waleko/SWE-bench-traceback · Datasets at Hugging Face

repo stringclasses 32 values	instance_id stringlengths 13 37	base_commit stringlengths 40 40	patch stringlengths 1 1.89M	problem_statement stringlengths 304 69k	hints_text stringlengths 0 246k	created_at stringlengths 20 20	FAIL_TO_PASS stringclasses 1 value	PASS_TO_PASS stringclasses 1 value	traceback stringlengths 64 23.4k	__index_level_0__ int64 29 19k
numpy/numpy	numpy__numpy-24299	b4c50e4862dd23ffcb27fe688d5e79e0456b2715	diff --git a/numpy/f2py/crackfortran.py b/numpy/f2py/crackfortran.py --- a/numpy/f2py/crackfortran.py +++ b/numpy/f2py/crackfortran.py @@ -1742,6 +1742,23 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: del d1[k] + if 'len' in d1 and 'array' in d1: + if d1['len'] == '': + d1['len'] = d1['array'] + del d1['array'] + elif typespec == 'character': + if ('charselector' not in edecl) or (not edecl['charselector']): + edecl['charselector'] = {} + if 'len' in edecl['charselector']: + del edecl['charselector']['len'] + edecl['charselector']['*'] = d1['len'] + del d1['len'] + else: + d1['array'] = d1['array'] + ',' + d1['len'] + del d1['len'] + errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( + typespec, e, typespec, ename, d1['array'])) + if 'len' in d1: if typespec in ['complex', 'integer', 'logical', 'real']: if ('kindselector' not in edecl) or (not edecl['kindselector']): @@ -1763,16 +1780,6 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: edecl['='] = d1['init'] - if 'len' in d1 and 'array' in d1: - if d1['len'] == '': - d1['len'] = d1['array'] - del d1['array'] - else: - d1['array'] = d1['array'] + ',' + d1['len'] - del d1['len'] - errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( - typespec, e, typespec, ename, d1['array'])) - if 'array' in d1: dm = 'dimension(%s)' % d1['array'] if 'attrspec' not in edecl or (not edecl['attrspec']):	BUG: f2py cannot compile files it used to be able to compile ### Describe the issue: I'm attempting to use f2py to compile some fortran code. I'm able to do this using the following numpy versions: - 1.21.0 - 1.22.0 - 1.23.0 - 1.24.0 - 1.24.2 Starting on v1.24.3, the same compilation code no longer works using f2py. Strangely enough, if I precompile one of files (LAPACK.f), the compilation can work with v1.24.3. ### Reproduce the code example: ```python # I don't see how I can produce runnable code using f2py, as it requires the fortran source code # This code produces a .so file in my project's home directory on numpy <= 1.24.2 but doesn't work on numpy >= 1.24.3 from pathlib import Path from numpy import f2py project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.f', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) # If I precompile LAPACK.f using: # /usr/bin/gfortran -Wall -g -ffixed-form -fno-second-underscore -g -fno-second-underscore -fPIC -O3 -funroll-loops -c LAPACK.f # then it works using numpy = 1.24.3. Note the only difference in the code is the LAPACK.f is now LAPACK.o project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.o', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) ``` ### Error message: ```shell There is no error message. f2py simply stops right before where it prints this line: INFO: compiling Fortran sources ``` ### Runtime information: Line 1 output: 1.24.0 3.10.6 (main, May 29 2023, 11:10:38) [GCC 11.3.0] (note that I have no idea why it says 1.24.0 when Pycharm assures me I'm using 1.24.2) Line 2 output: Exception ignored on calling ctypes callback function: <function ThreadpoolController._find_libraries_with_dl_iterate_phdr.<locals>.match_library_callback at 0x7f592f9c00d0> Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error [{'simd_extensions': {'baseline': ['SSE', 'SSE2', 'SSE3'], 'found': ['SSSE3', 'SSE41', 'POPCNT', 'SSE42', 'AVX', 'F16C', 'FMA3', 'AVX2'], 'not_found': ['AVX512F', 'AVX512CD', 'AVX512_KNL', 'AVX512_KNM', 'AVX512_SKX', 'AVX512_CLX', 'AVX512_CNL', 'AVX512_ICL']}}] None ### Context for the issue: I believe that LAPACK.f is one of the more widely used fortran codes. If it cannot compile in conjunction with other code, that could potentially disrupt a good number of users.	Ping @HaoZeke, also to confirm, can you check if the issue persists on 1.25.0? @seberg I can confirm that the issue persists on 1.25.0. It's actually why I noticed it in the first place. My code that used to run failed and then I tracked it down to the version number described above, in hopes that someone might have a better idea what caused it. On `1.24.4` the error is: ```bash ❯ f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f running build running config_cc INFO: unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc INFO: unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src INFO: build_src INFO: building extension "disort" sources INFO: f2py options: [] INFO: f2py:> /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c creating /tmp/tmpuul0531u/src.linux-x86_64-3.9 Reading fortran codes... Reading file 'BDREF.f' (format:fix,strict) Reading file 'DISOBRDF.f' (format:fix,strict) rmbadname1: Replacing "float" with "float_bn". rmbadname1: Replacing "len" with "len_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "float" with "float_bn". Reading file 'ERRPACK.f' (format:fix,strict) Reading file 'LAPACK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #745 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #1363 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #2115 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "char" with "char_bn". rmbadname1: Replacing "int" with "int_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #3999 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". Line #4273 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #5124 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #5462 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #6657 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Reading file 'LINPAK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Reading file 'RDI1MACH.f' (format:fix,strict) Post-processing... Block: disort Block: bdref Block: brdf_hapke Block: brdf_rpv Block: brdf_rossli Block: oceabrdf2 Block: shadow_eta Block: disobrdf {} In: :disort:DISOBRDF.f:surfac2 vars2fortran: No typespec for argument "nazz". Block: surfac2 Block: qgausn2 Block: zeroit2 Block: errmsg Block: wrtbad Block: wrtdim Block: tstbad Block: dgemm In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dger In: :disort:LAPACK.f:dger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetf2 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrf In: :disort:LAPACK.f:dgetrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrs In: :disort:LAPACK.f:dgetrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamch In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamc3 Block: dlaswp Block: dscal Block: dswap Block: dtrsm In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: idamax Block: ieeeck Block: ilaenv Block: iparmq In: :disort:LAPACK.f:iparmq get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: lsame Block: scopy Block: sgbtf2 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrf In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrs In: :disort:LAPACK.f:sgbtrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemm In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemv In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sger In: :disort:LAPACK.f:sger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: slaswp Block: stbsv In: :disort:LAPACK.f:stbsv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: strsm In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: xerbla Block: sgbco Block: sgbfa Block: sgbsl Block: sgeco Block: sgefa Block: sgesl Block: sasum Block: saxpy Block: sdot Block: sscal Block: sswap Block: isamax Block: r1mach Block: d1mach Block: i1mach Applying post-processing hooks... character_backward_compatibility_hook Post-processing (stage 2)... Building modules... Building module "disort"... Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "bdref"("bdref")... Constructing wrapper function "bdref"... bdref = bdref(mu,mup,dphi,brdf_type,brdf_arg) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_hapke"... brdf_hapke(mup,mu,dphi,b0,hh,w,pi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rpv"... brdf_rpv(mu_i,mu_r,dphi,rho0,kappa,g_hg,h0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rossli"... brdf_rossli(mu_i,mu_r,dphi,k_iso,k_vol,k_geo,alpha0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "oceabrdf2"... oceabrdf2(do_shadow,refrac_index,ws,mu_i,mu_r,dphi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "shadow_eta"("shadow_eta")... Constructing wrapper function "shadow_eta"... shadow_eta = shadow_eta(cos_theta,sigma_sq,pi) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "disobrdf"... rhoq,rhou,emust,bemst,bdr_beam_analytic = disobrdf(usrang,umu,fbeam,umu0,lamber,albedo,onlyfl,rhoq,rhou,emust,bemst,debug,phi,phi0,bdr_beam_analytic,brdf_type,brdf_arg,nmug,[nstr,numu,nphi]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "surfac2"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' surfac2(albedo,delm0,cmu,fbeam,lamber,mazim,onlyfl,pi,umu,umu0,usrang,bdr,emu,bem,rmu,rhoq,rhou,emust,bemst,debug,gmu,gwt,cosmp,brdf_type,brdf_arg,[mi,mxumu,nn,numu,nazz,nstr,nmug]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "qgausn2"... qgausn2(gmu,gwt,[m]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "zeroit2"... zeroit2(a,[length]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "errmsg"... errmsg(messag,fatal) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtbad"("wrtbad")... Constructing wrapper function "wrtbad"... wrtbad = wrtbad(varnam) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtdim"("wrtdim")... Constructing wrapper function "wrtdim"... wrtdim = wrtdim(dimnam,minval) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "tstbad"("tstbad")... Constructing wrapper function "tstbad"... tstbad = tstbad(varnam,relerr) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgemm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dger"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetf2"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgetf2(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrf"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgetrf(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrs"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgetrs(trans,n,nrhs,a,ipiv,b,info,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamch"("dlamch")... Constructing wrapper function "dlamch"... dlamch = dlamch(cmach) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamc3"("dlamc3")... Constructing wrapper function "dlamc3"... dlamc3 = dlamc3(a,b) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dlaswp"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dlaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dscal"... getarrdims:warning: assumed shape array, using 0 instead of '' dscal(n,da,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dswap"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dswap(n,dx,incx,dy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dtrsm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dtrsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "idamax"("idamax")... Constructing wrapper function "idamax"... getarrdims:warning: assumed shape array, using 0 instead of '' idamax = idamax(n,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ieeeck"("ieeeck")... Constructing wrapper function "ieeeck"... ieeeck = ieeeck(ispec,zero,one) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ilaenv"("ilaenv")... Constructing wrapper function "ilaenv"... ilaenv = ilaenv(ispec,name,opts,n1,n2,n3,n4) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} getarrdims:warning: assumed shape array, using 0 instead of '' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} iparmq = iparmq(ispec,name,opts,n,ilo,ihi,lwork) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "lsame"("lsame")... Constructing wrapper function "lsame"... lsame = lsame(ca,cb) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "scopy"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' scopy(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtf2"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbtf2(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrf"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbtrf(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrs"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbtrs(trans,n,kl,ku,nrhs,ab,ipiv,b,info,[ldab,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemv"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgemv(trans,m,n,alpha,a,x,incx,beta,y,incy,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sger"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "slaswp"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' slaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "stbsv"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' stbsv(uplo,trans,diag,n,k,a,x,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "strsm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' strsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "xerbla"... xerbla(srname,info) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbco"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbco(abd,n,ml,mu,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbfa"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbfa(abd,n,ml,mu,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbsl"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbsl(abd,n,ml,mu,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgeco"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgeco(a,n,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgefa"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgefa(a,n,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgesl"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgesl(a,n,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sasum"("sasum")... Constructing wrapper function "sasum"... getarrdims:warning: assumed shape array, using 0 instead of '' sasum = sasum(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "saxpy"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' saxpy(n,sa,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sdot"("sdot")... Constructing wrapper function "sdot"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sdot = sdot(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sscal"... getarrdims:warning: assumed shape array, using 0 instead of '' sscal(n,sa,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sswap"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sswap(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "isamax"("isamax")... Constructing wrapper function "isamax"... getarrdims:warning: assumed shape array, using 0 instead of '' isamax = isamax(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "r1mach"("r1mach")... Constructing wrapper function "r1mach"... r1mach = r1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "d1mach"("d1mach")... Constructing wrapper function "d1mach"... d1mach = d1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "i1mach"("i1mach")... Constructing wrapper function "i1mach"... i1mach = i1mach(i) Wrote C/API module "disort" to file "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c" Fortran 77 wrappers are saved to "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f" INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c' to sources. INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9' to include_dirs. copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.c -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.h -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f' to sources. INFO: build_src: building npy-pkg config files running build_ext INFO: customize UnixCCompiler INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmp27l9smg9/tmp creating /tmp/tmp27l9smg9/tmp/tmp27l9smg9 INFO: compile options: '-MMD -MF /tmp/tmp27l9smg9/file.c.d -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmp27l9smg9/file.c INFO: customize UnixCCompiler using build_ext INFO: get_default_fcompiler: matching types: '['arm', 'gnu95', 'intel', 'lahey', 'pg', 'nv', 'absoft', 'nag', 'vast', 'compaq', 'intele', 'intelem', 'gnu', 'g95', 'pathf95', 'nagfor', 'fujitsu']' INFO: customize ArmFlangCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-gfortran WARN: Could not locate executable armflang INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ar INFO: customize Gnu95FCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ld INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ranlib INFO: customize Gnu95FCompiler INFO: customize Gnu95FCompiler using build_ext INFO: building 'disort' extension INFO: compiling C sources INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmpuul0531u/tmp creating /tmp/tmpuul0531u/tmp/tmpuul0531u creating /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: compile options: '-DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 \| capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); \| ^ /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 \| capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); \| ^ error: Command "/home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC -DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c -o /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o -MMD -MF /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o.d" failed with exit status 1 ``` Where it seems that the relevant error is from changes to character handling: ```bash getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 \| capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); \| ^ /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 \| capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); \| ``` OTOH I would suspect something in the character handling of https://github.com/numpy/numpy/issues/23356 or https://github.com/numpy/numpy/pull/23194 Diffing the outputs b/w `1.24.3` and `1.24.4` only show: ```bash Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' ``` Indeed, `f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f skip: iparmq` does seem to compile as well. Will investigate ASAP. P.S. @kconnour the code being tested is [Pythonic-Distort](https://github.com/LDEO-CREW/Pythonic-DISORT/tree/main/disort4.0.99_f2py) right? Hi @HaoZeke, thanks for investigating! I'm really glad to see someone else is getting a similar error. Actually, the code I'm testing is on the api branch of [my repo](https://github.com/kconnour/pyRT_DISORT) but it should be extremely similar to the code in the repo you linked. We're both apparently trying to make a front-end to a popular open-source fortran algorithm. Note that I coded a workaround to this problem in pyproject.toml, where I force it to install numpy==1.24.0 in order to circumvent this issue... so if you ran the installation script from my repo, it shouldn't encounter this error. I'm happy to provide any additional info to help diagnose this issue! We ran into similar issues with our code, and after testing, it's because of having a decimal inside of a parameter declaration. Once that happens, everything later stops processing properly. The same seems be happening in the examples above If you look at the error log posted, https://github.com/numpy/numpy/issues/24008#issuecomment-1601586519, you'll see the first error message is the attached. ``` Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 ``` You could confirm whether this also works for your code by modifying that parameter to just be ONE=1, ZERO=0, and recompiling. If it gets past that point, that's the problem. We worked around this by reverting crackfortran.py to prior to the changes introduced here: https://github.com/numpy/numpy/pull/23637/files As to speculation: I think that the changes to the if blocks in https://github.com/numpy/numpy/pull/23637/commits caused an issue that assumes the input is an integer. Even more speculative - I see the kind selector is in that portion, and the kind selector requires integers. (https://numpy.org/doc/stable/f2py/advanced.html#dealing-with-kind-specifiers) . But I'm not sure - I didn't spend enough time on understanding crackfortran.py once we discovered that reverting crackfortran.py fixed our problem.	2023-07-31T06:34:32Z	[]	[]	Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error	10,207
numpy/numpy	numpy__numpy-24511	688f86133c78a3be569981a121d77166527d5b38	diff --git a/numpy/_build_utils/gitversion.py b/numpy/_build_utils/gitversion.py --- a/numpy/_build_utils/gitversion.py +++ b/numpy/_build_utils/gitversion.py @@ -24,6 +24,7 @@ def git_version(version): import subprocess import os.path + git_hash = '' try: p = subprocess.Popen( ['git', 'log', '-1', '--format="%H %aI"'], @@ -48,8 +49,6 @@ def git_version(version): # Only attach git tag to development versions if 'dev' in version: version += f'+git{git_date}.{git_hash[:7]}' - else: - git_hash = '' return version, git_hash	BUG: numpy 1.26.0b1 fails to build from sdist when no git is present ### Describe the issue: ---- `<edit>`: (@mattip) Adding the root cause: The version is missing from the sdist for 1.26b1, resulting in an attempt to get it via `git`. But that does not work because ... `<edit>` ---- The `gitversion.py` script introduced in #24196 is broken when no `git` binary is available. ### Reproduce the code example: ```python $ python3 numpy/_build_utils/gitversion.py ``` ### Error message: ```shell $ python3 numpy/_build_utils/gitversion.py Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value ``` ### Runtime information: N/A ### Context for the issue: https://github.com/sagemath/sage/pull/36123	Thanks for the report @mkoeppe. That's a bug indeed. The `gitversion.py` script should check for a file containing the needed git hash (as was done with `if fs.exists('_version_meson.py')` before gh-24196). Building from an sdist should work when `git` is not installed. @stefanv can you please have a look at this?	2023-08-23T13:02:33Z	[]	[]	Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value	10,212
numpy/numpy	numpy__numpy-24522	ef94040e9a7db18620d825782b8a14650f7c3346	diff --git a/numpy/_build_utils/gitversion.py b/numpy/_build_utils/gitversion.py --- a/numpy/_build_utils/gitversion.py +++ b/numpy/_build_utils/gitversion.py @@ -24,6 +24,7 @@ def git_version(version): import subprocess import os.path + git_hash = '' try: p = subprocess.Popen( ['git', 'log', '-1', '--format="%H %aI"'], @@ -48,8 +49,6 @@ def git_version(version): # Only attach git tag to development versions if 'dev' in version: version += f'+git{git_date}.{git_hash[:7]}' - else: - git_hash = '' return version, git_hash	BUG: numpy 1.26.0b1 fails to build from sdist when no git is present ### Describe the issue: ---- `<edit>`: (@mattip) Adding the root cause: The version is missing from the sdist for 1.26b1, resulting in an attempt to get it via `git`. But that does not work because ... `<edit>` ---- The `gitversion.py` script introduced in #24196 is broken when no `git` binary is available. ### Reproduce the code example: ```python $ python3 numpy/_build_utils/gitversion.py ``` ### Error message: ```shell $ python3 numpy/_build_utils/gitversion.py Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value ``` ### Runtime information: N/A ### Context for the issue: https://github.com/sagemath/sage/pull/36123	Thanks for the report @mkoeppe. That's a bug indeed. The `gitversion.py` script should check for a file containing the needed git hash (as was done with `if fs.exists('_version_meson.py')` before gh-24196). Building from an sdist should work when `git` is not installed. @stefanv can you please have a look at this?	2023-08-24T16:53:29Z	[]	[]	Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value	10,213
numpy/numpy	numpy__numpy-24542	9d64cc2bbe44292c46af768fb5da9b7968c2c6cb	diff --git a/numpy/f2py/crackfortran.py b/numpy/f2py/crackfortran.py --- a/numpy/f2py/crackfortran.py +++ b/numpy/f2py/crackfortran.py @@ -1742,6 +1742,23 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: del d1[k] + if 'len' in d1 and 'array' in d1: + if d1['len'] == '': + d1['len'] = d1['array'] + del d1['array'] + elif typespec == 'character': + if ('charselector' not in edecl) or (not edecl['charselector']): + edecl['charselector'] = {} + if 'len' in edecl['charselector']: + del edecl['charselector']['len'] + edecl['charselector']['*'] = d1['len'] + del d1['len'] + else: + d1['array'] = d1['array'] + ',' + d1['len'] + del d1['len'] + errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( + typespec, e, typespec, ename, d1['array'])) + if 'len' in d1: if typespec in ['complex', 'integer', 'logical', 'real']: if ('kindselector' not in edecl) or (not edecl['kindselector']): @@ -1763,16 +1780,6 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: edecl['='] = d1['init'] - if 'len' in d1 and 'array' in d1: - if d1['len'] == '': - d1['len'] = d1['array'] - del d1['array'] - else: - d1['array'] = d1['array'] + ',' + d1['len'] - del d1['len'] - errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( - typespec, e, typespec, ename, d1['array'])) - if 'array' in d1: dm = 'dimension(%s)' % d1['array'] if 'attrspec' not in edecl or (not edecl['attrspec']):	BUG: f2py cannot compile files it used to be able to compile ### Describe the issue: I'm attempting to use f2py to compile some fortran code. I'm able to do this using the following numpy versions: - 1.21.0 - 1.22.0 - 1.23.0 - 1.24.0 - 1.24.2 Starting on v1.24.3, the same compilation code no longer works using f2py. Strangely enough, if I precompile one of files (LAPACK.f), the compilation can work with v1.24.3. ### Reproduce the code example: ```python # I don't see how I can produce runnable code using f2py, as it requires the fortran source code # This code produces a .so file in my project's home directory on numpy <= 1.24.2 but doesn't work on numpy >= 1.24.3 from pathlib import Path from numpy import f2py project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.f', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) # If I precompile LAPACK.f using: # /usr/bin/gfortran -Wall -g -ffixed-form -fno-second-underscore -g -fno-second-underscore -fPIC -O3 -funroll-loops -c LAPACK.f # then it works using numpy = 1.24.3. Note the only difference in the code is the LAPACK.f is now LAPACK.o project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.o', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) ``` ### Error message: ```shell There is no error message. f2py simply stops right before where it prints this line: INFO: compiling Fortran sources ``` ### Runtime information: Line 1 output: 1.24.0 3.10.6 (main, May 29 2023, 11:10:38) [GCC 11.3.0] (note that I have no idea why it says 1.24.0 when Pycharm assures me I'm using 1.24.2) Line 2 output: Exception ignored on calling ctypes callback function: <function ThreadpoolController._find_libraries_with_dl_iterate_phdr.<locals>.match_library_callback at 0x7f592f9c00d0> Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error [{'simd_extensions': {'baseline': ['SSE', 'SSE2', 'SSE3'], 'found': ['SSSE3', 'SSE41', 'POPCNT', 'SSE42', 'AVX', 'F16C', 'FMA3', 'AVX2'], 'not_found': ['AVX512F', 'AVX512CD', 'AVX512_KNL', 'AVX512_KNM', 'AVX512_SKX', 'AVX512_CLX', 'AVX512_CNL', 'AVX512_ICL']}}] None ### Context for the issue: I believe that LAPACK.f is one of the more widely used fortran codes. If it cannot compile in conjunction with other code, that could potentially disrupt a good number of users.	Ping @HaoZeke, also to confirm, can you check if the issue persists on 1.25.0? @seberg I can confirm that the issue persists on 1.25.0. It's actually why I noticed it in the first place. My code that used to run failed and then I tracked it down to the version number described above, in hopes that someone might have a better idea what caused it. On `1.24.4` the error is: ```bash ❯ f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f running build running config_cc INFO: unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc INFO: unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src INFO: build_src INFO: building extension "disort" sources INFO: f2py options: [] INFO: f2py:> /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c creating /tmp/tmpuul0531u/src.linux-x86_64-3.9 Reading fortran codes... Reading file 'BDREF.f' (format:fix,strict) Reading file 'DISOBRDF.f' (format:fix,strict) rmbadname1: Replacing "float" with "float_bn". rmbadname1: Replacing "len" with "len_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "float" with "float_bn". Reading file 'ERRPACK.f' (format:fix,strict) Reading file 'LAPACK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #745 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #1363 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #2115 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "char" with "char_bn". rmbadname1: Replacing "int" with "int_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #3999 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". Line #4273 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #5124 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #5462 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #6657 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Reading file 'LINPAK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Reading file 'RDI1MACH.f' (format:fix,strict) Post-processing... Block: disort Block: bdref Block: brdf_hapke Block: brdf_rpv Block: brdf_rossli Block: oceabrdf2 Block: shadow_eta Block: disobrdf {} In: :disort:DISOBRDF.f:surfac2 vars2fortran: No typespec for argument "nazz". Block: surfac2 Block: qgausn2 Block: zeroit2 Block: errmsg Block: wrtbad Block: wrtdim Block: tstbad Block: dgemm In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dger In: :disort:LAPACK.f:dger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetf2 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrf In: :disort:LAPACK.f:dgetrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrs In: :disort:LAPACK.f:dgetrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamch In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamc3 Block: dlaswp Block: dscal Block: dswap Block: dtrsm In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: idamax Block: ieeeck Block: ilaenv Block: iparmq In: :disort:LAPACK.f:iparmq get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: lsame Block: scopy Block: sgbtf2 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrf In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrs In: :disort:LAPACK.f:sgbtrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemm In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemv In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sger In: :disort:LAPACK.f:sger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: slaswp Block: stbsv In: :disort:LAPACK.f:stbsv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: strsm In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: xerbla Block: sgbco Block: sgbfa Block: sgbsl Block: sgeco Block: sgefa Block: sgesl Block: sasum Block: saxpy Block: sdot Block: sscal Block: sswap Block: isamax Block: r1mach Block: d1mach Block: i1mach Applying post-processing hooks... character_backward_compatibility_hook Post-processing (stage 2)... Building modules... Building module "disort"... Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "bdref"("bdref")... Constructing wrapper function "bdref"... bdref = bdref(mu,mup,dphi,brdf_type,brdf_arg) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_hapke"... brdf_hapke(mup,mu,dphi,b0,hh,w,pi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rpv"... brdf_rpv(mu_i,mu_r,dphi,rho0,kappa,g_hg,h0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rossli"... brdf_rossli(mu_i,mu_r,dphi,k_iso,k_vol,k_geo,alpha0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "oceabrdf2"... oceabrdf2(do_shadow,refrac_index,ws,mu_i,mu_r,dphi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "shadow_eta"("shadow_eta")... Constructing wrapper function "shadow_eta"... shadow_eta = shadow_eta(cos_theta,sigma_sq,pi) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "disobrdf"... rhoq,rhou,emust,bemst,bdr_beam_analytic = disobrdf(usrang,umu,fbeam,umu0,lamber,albedo,onlyfl,rhoq,rhou,emust,bemst,debug,phi,phi0,bdr_beam_analytic,brdf_type,brdf_arg,nmug,[nstr,numu,nphi]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "surfac2"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' surfac2(albedo,delm0,cmu,fbeam,lamber,mazim,onlyfl,pi,umu,umu0,usrang,bdr,emu,bem,rmu,rhoq,rhou,emust,bemst,debug,gmu,gwt,cosmp,brdf_type,brdf_arg,[mi,mxumu,nn,numu,nazz,nstr,nmug]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "qgausn2"... qgausn2(gmu,gwt,[m]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "zeroit2"... zeroit2(a,[length]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "errmsg"... errmsg(messag,fatal) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtbad"("wrtbad")... Constructing wrapper function "wrtbad"... wrtbad = wrtbad(varnam) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtdim"("wrtdim")... Constructing wrapper function "wrtdim"... wrtdim = wrtdim(dimnam,minval) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "tstbad"("tstbad")... Constructing wrapper function "tstbad"... tstbad = tstbad(varnam,relerr) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgemm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dger"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetf2"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgetf2(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrf"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgetrf(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrs"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dgetrs(trans,n,nrhs,a,ipiv,b,info,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamch"("dlamch")... Constructing wrapper function "dlamch"... dlamch = dlamch(cmach) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamc3"("dlamc3")... Constructing wrapper function "dlamc3"... dlamc3 = dlamc3(a,b) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dlaswp"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dlaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dscal"... getarrdims:warning: assumed shape array, using 0 instead of '' dscal(n,da,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dswap"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dswap(n,dx,incx,dy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dtrsm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' dtrsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "idamax"("idamax")... Constructing wrapper function "idamax"... getarrdims:warning: assumed shape array, using 0 instead of '' idamax = idamax(n,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ieeeck"("ieeeck")... Constructing wrapper function "ieeeck"... ieeeck = ieeeck(ispec,zero,one) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ilaenv"("ilaenv")... Constructing wrapper function "ilaenv"... ilaenv = ilaenv(ispec,name,opts,n1,n2,n3,n4) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} getarrdims:warning: assumed shape array, using 0 instead of '' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} iparmq = iparmq(ispec,name,opts,n,ilo,ihi,lwork) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "lsame"("lsame")... Constructing wrapper function "lsame"... lsame = lsame(ca,cb) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "scopy"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' scopy(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtf2"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbtf2(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrf"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbtrf(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrs"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbtrs(trans,n,kl,ku,nrhs,ab,ipiv,b,info,[ldab,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemv"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgemv(trans,m,n,alpha,a,x,incx,beta,y,incy,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sger"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "slaswp"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' slaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "stbsv"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' stbsv(uplo,trans,diag,n,k,a,x,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "strsm"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' strsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "xerbla"... xerbla(srname,info) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbco"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbco(abd,n,ml,mu,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbfa"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbfa(abd,n,ml,mu,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbsl"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgbsl(abd,n,ml,mu,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgeco"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgeco(a,n,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgefa"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgefa(a,n,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgesl"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sgesl(a,n,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sasum"("sasum")... Constructing wrapper function "sasum"... getarrdims:warning: assumed shape array, using 0 instead of '' sasum = sasum(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "saxpy"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' saxpy(n,sa,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sdot"("sdot")... Constructing wrapper function "sdot"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sdot = sdot(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sscal"... getarrdims:warning: assumed shape array, using 0 instead of '' sscal(n,sa,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sswap"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' sswap(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "isamax"("isamax")... Constructing wrapper function "isamax"... getarrdims:warning: assumed shape array, using 0 instead of '' isamax = isamax(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "r1mach"("r1mach")... Constructing wrapper function "r1mach"... r1mach = r1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "d1mach"("d1mach")... Constructing wrapper function "d1mach"... d1mach = d1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "i1mach"("i1mach")... Constructing wrapper function "i1mach"... i1mach = i1mach(i) Wrote C/API module "disort" to file "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c" Fortran 77 wrappers are saved to "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f" INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c' to sources. INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9' to include_dirs. copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.c -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.h -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f' to sources. INFO: build_src: building npy-pkg config files running build_ext INFO: customize UnixCCompiler INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmp27l9smg9/tmp creating /tmp/tmp27l9smg9/tmp/tmp27l9smg9 INFO: compile options: '-MMD -MF /tmp/tmp27l9smg9/file.c.d -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmp27l9smg9/file.c INFO: customize UnixCCompiler using build_ext INFO: get_default_fcompiler: matching types: '['arm', 'gnu95', 'intel', 'lahey', 'pg', 'nv', 'absoft', 'nag', 'vast', 'compaq', 'intele', 'intelem', 'gnu', 'g95', 'pathf95', 'nagfor', 'fujitsu']' INFO: customize ArmFlangCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-gfortran WARN: Could not locate executable armflang INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ar INFO: customize Gnu95FCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ld INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ranlib INFO: customize Gnu95FCompiler INFO: customize Gnu95FCompiler using build_ext INFO: building 'disort' extension INFO: compiling C sources INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmpuul0531u/tmp creating /tmp/tmpuul0531u/tmp/tmpuul0531u creating /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: compile options: '-DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 \| capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); \| ^ /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 \| capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); \| ^ error: Command "/home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC -DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c -o /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o -MMD -MF /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o.d" failed with exit status 1 ``` Where it seems that the relevant error is from changes to character handling: ```bash getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'': ''}, 'attrspec': [], 'dimension': ['']} /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 \| capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); \| ^ /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 \| capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); \| ``` OTOH I would suspect something in the character handling of https://github.com/numpy/numpy/issues/23356 or https://github.com/numpy/numpy/pull/23194 Diffing the outputs b/w `1.24.3` and `1.24.4` only show: ```bash Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '' getarrdims:warning: assumed shape array, using 0 instead of '' ``` Indeed, `f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f skip: iparmq` does seem to compile as well. Will investigate ASAP. P.S. @kconnour the code being tested is [Pythonic-Distort](https://github.com/LDEO-CREW/Pythonic-DISORT/tree/main/disort4.0.99_f2py) right? Hi @HaoZeke, thanks for investigating! I'm really glad to see someone else is getting a similar error. Actually, the code I'm testing is on the api branch of [my repo](https://github.com/kconnour/pyRT_DISORT) but it should be extremely similar to the code in the repo you linked. We're both apparently trying to make a front-end to a popular open-source fortran algorithm. Note that I coded a workaround to this problem in pyproject.toml, where I force it to install numpy==1.24.0 in order to circumvent this issue... so if you ran the installation script from my repo, it shouldn't encounter this error. I'm happy to provide any additional info to help diagnose this issue! We ran into similar issues with our code, and after testing, it's because of having a decimal inside of a parameter declaration. Once that happens, everything later stops processing properly. The same seems be happening in the examples above If you look at the error log posted, https://github.com/numpy/numpy/issues/24008#issuecomment-1601586519, you'll see the first error message is the attached. ``` Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 ``` You could confirm whether this also works for your code by modifying that parameter to just be ONE=1, ZERO=0, and recompiling. If it gets past that point, that's the problem. We worked around this by reverting crackfortran.py to prior to the changes introduced here: https://github.com/numpy/numpy/pull/23637/files As to speculation: I think that the changes to the if blocks in https://github.com/numpy/numpy/pull/23637/commits caused an issue that assumes the input is an integer. Even more speculative - I see the kind selector is in that portion, and the kind selector requires integers. (https://numpy.org/doc/stable/f2py/advanced.html#dealing-with-kind-specifiers) . But I'm not sure - I didn't spend enough time on understanding crackfortran.py once we discovered that reverting crackfortran.py fixed our problem.	2023-08-25T14:39:05Z	[]	[]	Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error	10,215
numpy/numpy	numpy__numpy-258	b452014f0f9e8e6a69c6b95a62d70e3d99b9c0f9	diff --git a/numpy/distutils/ccompiler.py b/numpy/distutils/ccompiler.py --- a/numpy/distutils/ccompiler.py +++ b/numpy/distutils/ccompiler.py @@ -58,7 +58,11 @@ def CCompiler_spawn(self, cmd, display=None): if s: if is_sequence(cmd): cmd = ' '.join(list(cmd)) - print(o) + try: + print(o) + except UnicodeError: + # When installing through pip, `o` can contain non-ascii chars + pass if re.search('Too many open files', o): msg = '\nTry rerunning setup command until build succeeds.' else: diff --git a/numpy/distutils/command/build_clib.py b/numpy/distutils/command/build_clib.py --- a/numpy/distutils/command/build_clib.py +++ b/numpy/distutils/command/build_clib.py @@ -81,21 +81,23 @@ def run(self): if self.have_f_sources(): from numpy.distutils.fcompiler import new_fcompiler - self.fcompiler = new_fcompiler(compiler=self.fcompiler, - verbose=self.verbose, - dry_run=self.dry_run, - force=self.force, - requiref90='f90' in languages, - c_compiler=self.compiler) - if self.fcompiler is not None: - self.fcompiler.customize(self.distribution) + self._f_compiler = new_fcompiler(compiler=self.fcompiler, + verbose=self.verbose, + dry_run=self.dry_run, + force=self.force, + requiref90='f90' in languages, + c_compiler=self.compiler) + if self._f_compiler is not None: + self._f_compiler.customize(self.distribution) libraries = self.libraries self.libraries = None - self.fcompiler.customize_cmd(self) + self._f_compiler.customize_cmd(self) self.libraries = libraries - self.fcompiler.show_customization() + self._f_compiler.show_customization() + else: + self._f_compiler = None self.build_libraries(self.libraries) @@ -121,7 +123,7 @@ def build_libraries(self, libraries): def build_a_library(self, build_info, lib_name, libraries): # default compilers compiler = self.compiler - fcompiler = self.fcompiler + fcompiler = self._f_compiler sources = build_info.get('sources') if sources is None or not is_sequence(sources): @@ -233,7 +235,7 @@ def build_a_library(self, build_info, lib_name, libraries): debug=self.debug, extra_postargs=extra_postargs) - if requiref90 and self.fcompiler.module_dir_switch is None: + if requiref90 and self._f_compiler.module_dir_switch is None: # move new compiled F90 module files to module_build_dir for f in glob('.mod'): if f in existing_modules: diff --git a/numpy/distutils/fcompiler/hpux.py b/numpy/distutils/fcompiler/hpux.py --- a/numpy/distutils/fcompiler/hpux.py +++ b/numpy/distutils/fcompiler/hpux.py @@ -9,17 +9,17 @@ class HPUXFCompiler(FCompiler): version_pattern = r'HP F90 (?P<version>[^\s,])' executables = { - 'version_cmd' : ["<F90>", "+version"], + 'version_cmd' : ["f90", "+version"], 'compiler_f77' : ["f90"], 'compiler_fix' : ["f90"], 'compiler_f90' : ["f90"], - 'linker_so' : None, + 'linker_so' : ["ld", "-b"], 'archiver' : ["ar", "-cr"], 'ranlib' : ["ranlib"] } module_dir_switch = None #XXX: fix me module_include_switch = None #XXX: fix me - pic_flags = ['+pic=long'] + pic_flags = ['+Z'] def get_flags(self): return self.pic_flags + ['+ppu', '+DD64'] def get_flags_opt(self): diff --git a/numpy/distutils/fcompiler/ibm.py b/numpy/distutils/fcompiler/ibm.py --- a/numpy/distutils/fcompiler/ibm.py +++ b/numpy/distutils/fcompiler/ibm.py @@ -12,7 +12,7 @@ class IBMFCompiler(FCompiler): compiler_type = 'ibm' description = 'IBM XL Fortran Compiler' - version_pattern = r'(xlf\(1\)\s\|)IBM XL Fortran ((Advanced Edition \|)Version \|Enterprise Edition V)(?P<version>[^\s])' + version_pattern = r'(xlf\(1\)\s\|)IBM XL Fortran ((Advanced Edition \|)Version \|Enterprise Edition V\|for AIX, V)(?P<version>[^\s])' #IBM XL Fortran Enterprise Edition V10.1 for AIX \nVersion: 10.01.0000.0004 executables = { @@ -86,7 +86,7 @@ def get_flags_linker_so(self): return opt def get_flags_opt(self): - return ['-O5'] + return ['-O3'] if __name__ == '__main__': log.set_verbosity(2) diff --git a/numpy/distutils/fcompiler/pg.py b/numpy/distutils/fcompiler/pg.py --- a/numpy/distutils/fcompiler/pg.py +++ b/numpy/distutils/fcompiler/pg.py @@ -10,14 +10,14 @@ class PGroupFCompiler(FCompiler): compiler_type = 'pg' description = 'Portland Group Fortran Compiler' - version_pattern = r'\spg(f77\|f90\|hpf) (?P<version>[\d.-]+).' + version_pattern = r'\spg(f77\|f90\|hpf\|fortran) (?P<version>[\d.-]+).' if platform == 'darwin': executables = { - 'version_cmd' : ["<F77>", "-V 2>/dev/null"], - 'compiler_f77' : ["pgf77", "-dynamiclib"], - 'compiler_fix' : ["pgf90", "-Mfixed", "-dynamiclib"], - 'compiler_f90' : ["pgf90", "-dynamiclib"], + 'version_cmd' : ["<F77>", "-V"], + 'compiler_f77' : ["pgfortran", "-dynamiclib"], + 'compiler_fix' : ["pgfortran", "-Mfixed", "-dynamiclib"], + 'compiler_f90' : ["pgfortran", "-dynamiclib"], 'linker_so' : ["libtool"], 'archiver' : ["ar", "-cr"], 'ranlib' : ["ranlib"] @@ -25,11 +25,11 @@ class PGroupFCompiler(FCompiler): pic_flags = [''] else: executables = { - 'version_cmd' : ["<F77>", "-V 2>/dev/null"], - 'compiler_f77' : ["pgf77"], - 'compiler_fix' : ["pgf90", "-Mfixed"], - 'compiler_f90' : ["pgf90"], - 'linker_so' : ["pgf90","-shared","-fpic"], + 'version_cmd' : ["<F77>", "-V"], + 'compiler_f77' : ["pgfortran"], + 'compiler_fix' : ["pgfortran", "-Mfixed"], + 'compiler_f90' : ["pgfortran"], + 'linker_so' : ["pgfortran","-shared","-fpic"], 'archiver' : ["ar", "-cr"], 'ranlib' : ["ranlib"] } diff --git a/numpy/distutils/system_info.py b/numpy/distutils/system_info.py --- a/numpy/distutils/system_info.py +++ b/numpy/distutils/system_info.py @@ -125,6 +125,7 @@ from distutils.dist import Distribution import distutils.sysconfig from distutils import log +from distutils.util import get_platform from numpy.distutils.exec_command import \ find_executable, exec_command, get_pythonexe @@ -193,14 +194,23 @@ def libpaths(paths,bits): '/opt/local/lib','/sw/lib'], platform_bits) default_include_dirs = ['/usr/local/include', '/opt/include', '/usr/include', + # path of umfpack under macports + '/opt/local/include/ufsparse', '/opt/local/include', '/sw/include', '/usr/include/suitesparse'] default_src_dirs = ['.','/usr/local/src', '/opt/src','/sw/src'] - default_x11_lib_dirs = libpaths(['/usr/X11R6/lib','/usr/X11/lib', '/usr/lib'], platform_bits) default_x11_include_dirs = ['/usr/X11R6/include','/usr/X11/include', '/usr/include'] + if os.path.exists('/usr/lib/X11'): + globbed_x11_dir = glob('/usr/lib//libX11.so') + if globbed_x11_dir: + x11_so_dir = os.path.split(globbed_x11_dir[0])[0] + default_x11_lib_dirs.extend([x11_so_dir, '/usr/lib/X11']) + default_x11_include_dirs.extend(['/usr/lib/X11/include', + '/usr/include/X11']) + if os.path.join(sys.prefix, 'lib') not in default_lib_dirs: default_lib_dirs.insert(0,os.path.join(sys.prefix, 'lib')) @@ -1273,7 +1283,6 @@ def get_atlas_version(**config): result = _cached_atlas_version[key] = atlas_version, info return result -from distutils.util import get_platform class lapack_opt_info(system_info): @@ -1284,7 +1293,8 @@ def calc_info(self): if sys.platform=='darwin' and not os.environ.get('ATLAS',None): args = [] link_args = [] - if get_platform()[-4:] == 'i386': + if get_platform()[-4:] == 'i386' or 'intel' in get_platform() or \ + 'i386' in platform.platform(): intel = 1 else: intel = 0 @@ -1371,7 +1381,8 @@ def calc_info(self): if sys.platform=='darwin' and not os.environ.get('ATLAS',None): args = [] link_args = [] - if get_platform()[-4:] == 'i386': + if get_platform()[-4:] == 'i386' or 'intel' in get_platform() or \ + 'i386' in platform.platform(): intel = 1 else: intel = 0 diff --git a/numpy/distutils/unixccompiler.py b/numpy/distutils/unixccompiler.py --- a/numpy/distutils/unixccompiler.py +++ b/numpy/distutils/unixccompiler.py @@ -17,6 +17,18 @@ # Note that UnixCCompiler._compile appeared in Python 2.3 def UnixCCompiler__compile(self, obj, src, ext, cc_args, extra_postargs, pp_opts): """Compile a single source files with a Unix-style compiler.""" + # HP ad-hoc fix, see ticket 1383 + ccomp = self.compiler_so + if ccomp[0] == 'aCC': + # remove flags that will trigger ANSI-C mode for aCC + if '-Ae' in ccomp: + ccomp.remove('-Ae') + if '-Aa' in ccomp: + ccomp.remove('-Aa') + # add flags for (almost) sane C++ handling + ccomp += ['-AA'] + self.compiler_so = ccomp + display = '%s: %s' % (os.path.basename(self.compiler_so[0]),src) try: self.spawn(self.compiler_so + cc_args + [src, '-o', obj] + diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -114,7 +114,12 @@ def write_version_py(filename='numpy/version.py'): GIT_REVISION = git_version() elif os.path.exists('numpy/version.py'): # must be a source distribution, use existing version file - from numpy.version import git_revision as GIT_REVISION + try: + from numpy.version import git_revision as GIT_REVISION + except ImportError: + raise ImportError("Unable to import git_revision. Try removing " \ + "numpy/version.py and the build directory " \ + "before building.") else: GIT_REVISION = "Unknown" @@ -162,6 +167,19 @@ def setup_package(): if os.path.isfile(site_cfg): shutil.copy(site_cfg, src_path) + # Ugly hack to make pip work with Python 3, see #1857. + # Explanation: pip messes with __file__ which interacts badly with the + # change in directory due to the 2to3 conversion. Therefore we restore + # __file__ to what it would have been otherwise. + global __file__ + __file__ = os.path.join(os.curdir, os.path.basename(__file__)) + if '--egg-base' in sys.argv: + # Change pip-egg-info entry to absolute path, so pip can find it + # after changing directory. + idx = sys.argv.index('--egg-base') + if sys.argv[idx + 1] == 'pip-egg-info': + sys.argv[idx + 1] = os.path.join(local_path, 'pip-egg-info') + old_path = os.getcwd() os.chdir(src_path) sys.path.insert(0, src_path)	Error in linalg.norm() (Trac #785) _Original ticket http://projects.scipy.org/numpy/ticket/785 on 2008-05-09 by trac user nick, assigned to unknown._ While working on a unit test for linalg.norm() (see ticket #1361), I discovered that if a vector is passed into the norm() method with 'fro' as the argument for the ord, an error occurs. Example: ``` >>> from numpy import linalg >>> a = [1,2,3,4] >>> linalg.norm(a,'fro') Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/tmp/lib/python2.5/site-packages/numpy/linalg/linalg.py", line 1262, in norm return ((abs(x)ord).sum())(1.0/ord) TypeError: unsupported operand type(s) for ** or pow(): 'numpy.ndarray' and 'str' >>> ``` Performing the same test but omiting the parameter 'fro' allows the method to execute normally. A test case that exposes this is available in ticket #1361.		2012-04-22T11:18:04Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/tmp/lib/python2.5/site-packages/numpy/linalg/linalg.py", line 1262, in norm return ((abs(x)ord).sum())(1.0/ord) TypeError: unsupported operand type(s) for ** or pow(): 'numpy.ndarray' and 'str'	10,219
numpy/numpy	numpy__numpy-3447	2a5c2c8227b600654f31ed346c73cce77bef554d	diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -5643,6 +5643,9 @@ def __iter__(self): else: yield d + def __len__(self): + return self._data.__len__() + def filled(self, fill_value=None): """ Return a copy with masked fields filled with a given value.	MaskedArray record: TypeError: len() of unsized object (Trac #2116) _Original ticket http://projects.scipy.org/numpy/ticket/2116 on 2012-04-26 by trac user mwtoews, assigned to @pierregm._ I'm getting a inconsistent error while trying to get the length of a record from a masked array, but only when the mask is enabled. See the example: ``` import numpy as np my_dtype = [('a','i'),('b','f')] mar = np.ma.zeros(5, my_dtype) # Length of the first record; no error raised assert len(mar[0]) == 2 # Change mask for one of the fields print(mar[0]) # (0, 0.0) mar.mask[0][0] = True print(mar[0]) # (--, 0.0) # Repeat same command as above to reveal this bug assert len(mar[0]) == 2 ``` Raises the error: ``` Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object ``` I am using !NumPy version 1.6.1 obtained from http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy and the version of Python is 2.5.1 [MSC v.1310 32 bit (Intel)].		2013-06-16T01:41:13Z	[]	[]	Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object	10,263
numpy/numpy	numpy__numpy-3452	5ba9fead691439697c6d1c3768d4205cbe822bc2	diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -5593,6 +5593,9 @@ def __iter__(self): else: yield d + def __len__(self): + return self._data.__len__() + def filled(self, fill_value=None): """ Return a copy with masked fields filled with a given value. @@ -5930,9 +5933,10 @@ class _frommethod: Name of the method to transform. """ - def __init__(self, methodname): + def __init__(self, methodname, reversed=False): self.__name__ = methodname self.__doc__ = self.getdoc() + self.reversed = reversed # def getdoc(self): "Return the doc of the function (from the doc of the method)." @@ -5944,6 +5948,11 @@ def getdoc(self): return doc # def __call__(self, a, args, params): + if self.reversed: + args = list(args) + arr = args[0] + args[0] = a + a = arr # Get the method from the array (if possible) method_name = self.__name__ method = getattr(a, method_name, None) @@ -5960,7 +5969,7 @@ def __call__(self, a, args, **params): all = _frommethod('all') anomalies = anom = _frommethod('anom') any = _frommethod('any') -compress = _frommethod('compress') +compress = _frommethod('compress', reversed=True) cumprod = _frommethod('cumprod') cumsum = _frommethod('cumsum') copy = _frommethod('copy')	MaskedArray record: TypeError: len() of unsized object (Trac #2116) _Original ticket http://projects.scipy.org/numpy/ticket/2116 on 2012-04-26 by trac user mwtoews, assigned to @pierregm._ I'm getting a inconsistent error while trying to get the length of a record from a masked array, but only when the mask is enabled. See the example: ``` import numpy as np my_dtype = [('a','i'),('b','f')] mar = np.ma.zeros(5, my_dtype) # Length of the first record; no error raised assert len(mar[0]) == 2 # Change mask for one of the fields print(mar[0]) # (0, 0.0) mar.mask[0][0] = True print(mar[0]) # (--, 0.0) # Repeat same command as above to reveal this bug assert len(mar[0]) == 2 ``` Raises the error: ``` Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object ``` I am using !NumPy version 1.6.1 obtained from http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy and the version of Python is 2.5.1 [MSC v.1310 32 bit (Intel)].		2013-06-17T01:40:29Z	[]	[]	Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object	10,267
numpy/numpy	numpy__numpy-3854	106306dc9b0e3aaf0543d78c7c20761ef03f0213	diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -368,10 +368,11 @@ def solve(a, b): gufunc = _umath_linalg.solve1 else: - if a.shape[-1] == 0 and b.shape[-2] == 0: - a = a.reshape(a.shape[:-1] + (1,)) - bc = broadcast(a, b) - return wrap(empty(bc.shape, dtype=result_t)) + if b.size == 0: + if (a.shape[-1] == 0 and b.shape[-2] == 0) or b.shape[-1] == 0: + a = a[:,:1].reshape(a.shape[:-1] + (1,)) + bc = broadcast(a, b) + return wrap(empty(bc.shape, dtype=result_t)) gufunc = _umath_linalg.solve	solve() fails on 0-sized axis With current 1.8, 5a0d09c: ``` import numpy as np np.linalg.solve(np.eye(3), np.zeros((3, 0))) ``` gives ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/njs/src/numpy/.tox/py27/local/lib/python2.7/site-packages/numpy/linalg/linalg.py", line 380, in solve r = gufunc(a, b, signature=signature, extobj=extobj) ValueError: cannot remove a zero-sized axis from an iterator ``` The result should be an array with shape `(3, 0)`. This is a regression from 1.7, hence marking as a blocker. Thanks to Jens Jørgen Mortensen for the original report: http://mail.scipy.org/pipermail/numpy-discussion/2013-October/067898.html		2013-10-02T16:46:10Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/njs/src/numpy/.tox/py27/local/lib/python2.7/site-packages/numpy/linalg/linalg.py", line 380, in solve r = gufunc(a, b, signature=signature, extobj=extobj) ValueError: cannot remove a zero-sized axis from an iterator	10,284
numpy/numpy	numpy__numpy-4276	bd5894b29b897f16da8a3d64e0df94e93d6b2d4a	diff --git a/numpy/core/_methods.py b/numpy/core/_methods.py --- a/numpy/core/_methods.py +++ b/numpy/core/_methods.py @@ -63,8 +63,10 @@ def _mean(a, axis=None, dtype=None, out=None, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -107,8 +109,10 @@ def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -118,7 +122,9 @@ def _std(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.sqrt(ret, out=ret) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(um.sqrt(ret)) + else: + ret = um.sqrt(ret) return ret	mean() (and median()) should work with "object" arrays With NumPy 1.8, `mean()` started to break when calculating the (global) mean of an array that contains objects (arrays with an object `dtype`). This also breaks `median()` on such arrays. Here is an example: ``` >>> numpy.arange(10).astype(object).mean() Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype' ``` Another example is case of numbers with uncertainties from the uncertainties package (https://github.com/lebigot/uncertainties/issues/22). I think that it would be better if NumPy did not assume that scalar results have a `dtype`, since arrays can contain objects that have a meaningful mean. I believe that such objects should not be forced to have a `dtype`, which is obviously NumPy specific (they even can't, for Python scalars like floats). Furthermore, a `dtype` is in principle not necessary for the calculation of the mean of such objects, so it would look strange if they had to have one. The problem is that `numpy.mean()` assumes that the intermediate result obtained has a `dtype` (with a `type` attribute). Therefore, I suggest that NumPy's `mean()` also handle arrays of objects that are not of the standard NumPy types (their `dtype` is object, and they contain objects that have a meaningful mean, like ints, floats, numbers with uncertainties, etc.).	introduced in f16b12e87 by @charris Hmmm, this is annoying. It seems to me like the only way to fix this is probably to see if `dtype` is given, and then use `np.dtype(dtype).type(...)` and otherwise just do the plain operation? Yeah, I was thinking along the same lines. I would be curious to see what the issue that prompted the change was, to see if I can come up with any kind of better suggestion (who knows): what was the problem, exactly? The scalar returns didn't preserve type, i.e., float32 would go to float64. That was on account of type precedence between scalars being different than type precedence between scalars and arrays. @charris do we care even about that? Or is it enough if the passed in dtype actually gets honored? I honestly have troubles to figure out a good method of preserving the type quite right for the scalar result. I now think we may have to just check for object dtype input (or passed in dtype). The most secure method I can think of would be a new keyword argument to the ufuncs to skip PyArray_Return (would probably be slower though), but unless that is useful elsewhere it is not worth the trouble either. Any progress on this one? It costed me a hour of debugging today. If doing this properly is hard, please consider fixing the error message so it is obvious what's wrong. @charris do you have time to have a look at this? I also think we accumulated enough fixes to warrant a 1.8.1 release if we add this and the C99 windows fix. thoughts? I'll get it done today sometime. Agree on 1.8.1, I came to that conclusion this morning. We should also fix the `divide` and `true_divide` ufuncs when the `dtype` is given. I need to think about this a bit more before putting up a fix.	2014-02-10T05:05:28Z	[]	[]	Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype'	10,290
numpy/numpy	numpy__numpy-4297	95f7a469b1e9ce460e31c41e1bd897ceff396f6b	diff --git a/numpy/core/_methods.py b/numpy/core/_methods.py --- a/numpy/core/_methods.py +++ b/numpy/core/_methods.py @@ -63,8 +63,10 @@ def _mean(a, axis=None, dtype=None, out=None, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -107,8 +109,10 @@ def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -118,7 +122,9 @@ def _std(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.sqrt(ret, out=ret) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(um.sqrt(ret)) + else: + ret = um.sqrt(ret) return ret diff --git a/numpy/distutils/fcompiler/gnu.py b/numpy/distutils/fcompiler/gnu.py --- a/numpy/distutils/fcompiler/gnu.py +++ b/numpy/distutils/fcompiler/gnu.py @@ -35,13 +35,13 @@ class GnuFCompiler(FCompiler): def gnu_version_match(self, version_string): """Handle the different versions of GNU fortran compilers""" - m = re.match(r'GNU Fortran', version_string) + m = re.search(r'GNU Fortran', version_string) if not m: return None - m = re.match(r'GNU Fortran\s+95.?([0-9-.]+)', version_string) + m = re.search(r'GNU Fortran\s+95.?([0-9-.]+)', version_string) if m: return ('gfortran', m.group(1)) - m = re.match(r'GNU Fortran.?\-?([0-9-.]+)', version_string) + m = re.search(r'GNU Fortran.?\-?([0-9-.]+)', version_string) if m: v = m.group(1) if v.startswith('0') or v.startswith('2') or v.startswith('3'): diff --git a/numpy/f2py/__init__.py b/numpy/f2py/__init__.py --- a/numpy/f2py/__init__.py +++ b/numpy/f2py/__init__.py @@ -28,20 +28,20 @@ def compile(source, from numpy.distutils.exec_command import exec_command import tempfile if source_fn is None: - fname = os.path.join(tempfile.mktemp()+'.f') + f = tempfile.NamedTemporaryFile(suffix='.f') else: - fname = source_fn - - f = open(fname, 'w') - f.write(source) - f.close() - - args = ' -c -m %s %s %s'%(modulename, fname, extra_args) - c = '%s -c "import numpy.f2py as f2py2e;f2py2e.main()" %s' %(sys.executable, args) - s, o = exec_command(c) - if source_fn is None: - try: os.remove(fname) - except OSError: pass + f = open(source_fn, 'w') + + try: + f.write(source) + f.flush() + + args = ' -c -m %s %s %s'%(modulename, f.name, extra_args) + c = '%s -c "import numpy.f2py as f2py2e;f2py2e.main()" %s' % \ + (sys.executable, args) + s, o = exec_command(c) + finally: + f.close() return s from numpy.testing import Tester diff --git a/numpy/f2py/f2py2e.py b/numpy/f2py/f2py2e.py --- a/numpy/f2py/f2py2e.py +++ b/numpy/f2py/f2py2e.py @@ -91,7 +91,7 @@ --lower is assumed with -h key, and --no-lower without -h key. --build-dir <dirname> All f2py generated files are created in <dirname>. - Default is tempfile.mktemp(). + Default is tempfile.mkdtemp(). --overwrite-signature Overwrite existing signature file. @@ -428,7 +428,7 @@ def run_compile(): del sys.argv[i] else: remove_build_dir = 1 - build_dir = os.path.join(tempfile.mktemp()) + build_dir = tempfile.mkdtemp() _reg1 = re.compile(r'[-][-]link[-]') sysinfo_flags = [_m for _m in sys.argv[1:] if _reg1.match(_m)] diff --git a/numpy/lib/financial.py b/numpy/lib/financial.py --- a/numpy/lib/financial.py +++ b/numpy/lib/financial.py @@ -628,21 +628,29 @@ def irr(values): Examples -------- - >>> print round(np.irr([-100, 39, 59, 55, 20]), 5) + >>> round(irr([-100, 39, 59, 55, 20]), 5) 0.28095 + >>> round(irr([-100, 0, 0, 74]), 5) + -0.0955 + >>> round(irr([-100, 100, 0, -7]), 5) + -0.0833 + >>> round(irr([-100, 100, 0, 7]), 5) + 0.06206 + >>> round(irr([-5, 10.5, 1, -8, 1]), 5) + 0.0886 (Compare with the Example given for numpy.lib.financial.npv) """ res = np.roots(values[::-1]) - # Find the root(s) between 0 and 1 - mask = (res.imag == 0) & (res.real > 0) & (res.real <= 1) - res = res[mask].real + mask = (res.imag == 0) & (res.real > 0) if res.size == 0: return np.nan + res = res[mask].real + # NPV(rate) = 0 can have more than one solution so we return + # only the solution closest to zero. rate = 1.0/res - 1 - if rate.size == 1: - rate = rate.item() + rate = rate.item(np.argmin(np.abs(rate))) return rate def npv(rate, values): diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -18,11 +18,10 @@ DOCLINES = __doc__.split("\n") import os -import shutil import sys -import re import subprocess + if sys.version_info[:2] < (2, 6) or (3, 0) <= sys.version_info[0:2] < (3, 2): raise RuntimeError("Python version 2.6, 2.7 or >= 3.2 required.") @@ -31,6 +30,7 @@ else: import __builtin__ as builtins + CLASSIFIERS = """\ Development Status :: 5 - Production/Stable Intended Audience :: Science/Research @@ -47,24 +47,13 @@ Operating System :: MacOS """ -NAME = 'numpy' -MAINTAINER = "NumPy Developers" -MAINTAINER_EMAIL = "numpy-discussion@scipy.org" -DESCRIPTION = DOCLINES[0] -LONG_DESCRIPTION = "\n".join(DOCLINES[2:]) -URL = "http://www.numpy.org" -DOWNLOAD_URL = "http://sourceforge.net/projects/numpy/files/NumPy/" -LICENSE = 'BSD' -CLASSIFIERS = [_f for _f in CLASSIFIERS.split('\n') if _f] -AUTHOR = "Travis E. Oliphant et al." -AUTHOR_EMAIL = "oliphant@enthought.com" -PLATFORMS = ["Windows", "Linux", "Solaris", "Mac OS-X", "Unix"] MAJOR = 1 MINOR = 8 MICRO = 0 ISRELEASED = False VERSION = '%d.%d.%d' % (MAJOR, MINOR, MICRO) + # Return the git revision as a string def git_version(): def _minimal_ext_cmd(cmd): @@ -100,18 +89,7 @@ def _minimal_ext_cmd(cmd): builtins.__NUMPY_SETUP__ = True -def write_version_py(filename='numpy/version.py'): - cnt = """ -# THIS FILE IS GENERATED FROM NUMPY SETUP.PY -short_version = '%(version)s' -version = '%(version)s' -full_version = '%(full_version)s' -git_revision = '%(git_revision)s' -release = %(isrelease)s - -if not release: - version = full_version -""" +def get_version_info(): # Adding the git rev number needs to be done inside write_version_py(), # otherwise the import of numpy.version messes up the build under Python 3. FULLVERSION = VERSION @@ -131,6 +109,23 @@ def write_version_py(filename='numpy/version.py'): if not ISRELEASED: FULLVERSION += '.dev-' + GIT_REVISION[:7] + return FULLVERSION, GIT_REVISION + + +def write_version_py(filename='numpy/version.py'): + cnt = """ +# THIS FILE IS GENERATED FROM NUMPY SETUP.PY +short_version = '%(version)s' +version = '%(version)s' +full_version = '%(full_version)s' +git_revision = '%(git_revision)s' +release = %(isrelease)s + +if not release: + version = full_version +""" + FULLVERSION, GIT_REVISION = get_version_info() + a = open(filename, 'w') try: a.write(cnt % {'version': VERSION, @@ -140,6 +135,7 @@ def write_version_py(filename='numpy/version.py'): finally: a.close() + def configuration(parent_package='',top_path=None): from numpy.distutils.misc_util import Configuration @@ -155,8 +151,36 @@ def configuration(parent_package='',top_path=None): return config -def setup_package(): +def check_submodules(): + """ verify that the submodules are checked out and clean + use `git submodule update --init`; on failure + """ + if not os.path.exists('.git'): + return + with open('.gitmodules') as f: + for l in f: + if 'path' in l: + p = l.split('=')[-1].strip() + if not os.path.exists(p): + raise ValueError('Submodule %s missing' % p) + + + proc = subprocess.Popen(['git', 'submodule', 'status'], + stdout=subprocess.PIPE) + status, _ = proc.communicate() + status = status.decode("ascii", "replace") + for line in status.splitlines(): + if line.startswith('-') or line.startswith('+'): + raise ValueError('Submodule not clean: %s' % line) + +from distutils.command.sdist import sdist +class sdist_checked(sdist): + """ check submodules on sdist to prevent incomplete tarballs """ + def run(self): + check_submodules() + sdist.run(self) +def setup_package(): src_path = os.path.dirname(os.path.abspath(sys.argv[0])) old_path = os.getcwd() os.chdir(src_path) @@ -165,28 +189,51 @@ def setup_package(): # Rewrite the version file everytime write_version_py() + metadata = dict( + name = 'numpy', + maintainer = "NumPy Developers", + maintainer_email = "numpy-discussion@scipy.org", + description = DOCLINES[0], + long_description = "\n".join(DOCLINES[2:]), + url = "http://www.numpy.org", + author = "Travis E. Oliphant et al.", + download_url = "http://sourceforge.net/projects/numpy/files/NumPy/", + license = 'BSD', + classifiers=[_f for _f in CLASSIFIERS.split('\n') if _f], + platforms = ["Windows", "Linux", "Solaris", "Mac OS-X", "Unix"], + test_suite='nose.collector', + cmdclass={"sdist": sdist_checked}, + ) + # Run build - from numpy.distutils.core import setup + if len(sys.argv) >= 2 and ('--help' in sys.argv[1:] or + sys.argv[1] in ('--help-commands', 'egg_info', '--version', + 'clean')): + # Use setuptools for these commands (they don't work well or at all + # with distutils). For normal builds use distutils. + try: + from setuptools import setup + except ImportError: + from distutils.core import setup + + FULLVERSION, GIT_REVISION = get_version_info() + metadata['version'] = FULLVERSION + elif len(sys.argv) >= 2 and sys.argv[1] == 'bdist_wheel': + # bdist_wheel needs setuptools + import setuptools + from numpy.distutils.core import setup + metadata['configuration'] = configuration + else: + from numpy.distutils.core import setup + metadata['configuration'] = configuration try: - setup( - name=NAME, - maintainer=MAINTAINER, - maintainer_email=MAINTAINER_EMAIL, - description=DESCRIPTION, - long_description=LONG_DESCRIPTION, - url=URL, - download_url=DOWNLOAD_URL, - license=LICENSE, - classifiers=CLASSIFIERS, - author=AUTHOR, - author_email=AUTHOR_EMAIL, - platforms=PLATFORMS, - configuration=configuration ) + setup(**metadata) finally: del sys.path[0] os.chdir(old_path) return + if __name__ == '__main__': setup_package()	mean() (and median()) should work with "object" arrays With NumPy 1.8, `mean()` started to break when calculating the (global) mean of an array that contains objects (arrays with an object `dtype`). This also breaks `median()` on such arrays. Here is an example: ``` >>> numpy.arange(10).astype(object).mean() Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype' ``` Another example is case of numbers with uncertainties from the uncertainties package (https://github.com/lebigot/uncertainties/issues/22). I think that it would be better if NumPy did not assume that scalar results have a `dtype`, since arrays can contain objects that have a meaningful mean. I believe that such objects should not be forced to have a `dtype`, which is obviously NumPy specific (they even can't, for Python scalars like floats). Furthermore, a `dtype` is in principle not necessary for the calculation of the mean of such objects, so it would look strange if they had to have one. The problem is that `numpy.mean()` assumes that the intermediate result obtained has a `dtype` (with a `type` attribute). Therefore, I suggest that NumPy's `mean()` also handle arrays of objects that are not of the standard NumPy types (their `dtype` is object, and they contain objects that have a meaningful mean, like ints, floats, numbers with uncertainties, etc.).	introduced in f16b12e87 by @charris Hmmm, this is annoying. It seems to me like the only way to fix this is probably to see if `dtype` is given, and then use `np.dtype(dtype).type(...)` and otherwise just do the plain operation? Yeah, I was thinking along the same lines. I would be curious to see what the issue that prompted the change was, to see if I can come up with any kind of better suggestion (who knows): what was the problem, exactly? The scalar returns didn't preserve type, i.e., float32 would go to float64. That was on account of type precedence between scalars being different than type precedence between scalars and arrays. @charris do we care even about that? Or is it enough if the passed in dtype actually gets honored? I honestly have troubles to figure out a good method of preserving the type quite right for the scalar result. I now think we may have to just check for object dtype input (or passed in dtype). The most secure method I can think of would be a new keyword argument to the ufuncs to skip PyArray_Return (would probably be slower though), but unless that is useful elsewhere it is not worth the trouble either.	2014-02-15T17:25:25Z	[]	[]	Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype'	10,291
numpy/numpy	numpy__numpy-4304	68ae04741f73953ef5680cb80fbb3dde5f160c90	diff --git a/numpy/lib/twodim_base.py b/numpy/lib/twodim_base.py --- a/numpy/lib/twodim_base.py +++ b/numpy/lib/twodim_base.py @@ -25,7 +25,7 @@ def fliplr(m): Parameters ---------- m : array_like - Input array. + Input array, must be at least 2-D. Returns ------- @@ -40,8 +40,7 @@ def fliplr(m): Notes ----- - Equivalent to A[:,::-1]. Does not require the array to be - two-dimensional. + Equivalent to A[:,::-1]. Requires the array to be at least 2-D. Examples --------	fliplr documentation incorrectly states 2-d not required [`fliplr` documentation](http://docs.scipy.org/doc/numpy/reference/generated/numpy.fliplr.html) incorrectly states 2-d _not_ required, but that is _not_ true. I think it is a typo, since `fliplr` was probably copied from `flipud` which _really_ doesn't require a 2-d array. ``` >>> import numpy as np >>> a = np.array([1,2,3,4]) >>> np.fliplr(a) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Python27\lib\site-packages\numpy\lib\twodim_base.py", line 61, in fliplr raise ValueError("Input must be >= 2-d.") ValueError: Input must be >= 2-d. ```		2014-02-16T20:32:54Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Python27\lib\site-packages\numpy\lib\twodim_base.py", line 61, in fliplr raise ValueError("Input must be >= 2-d.") ValueError: Input must be >= 2-d.	10,293
numpy/numpy	numpy__numpy-4305	68ae04741f73953ef5680cb80fbb3dde5f160c90	diff --git a/numpy/f2py/cfuncs.py b/numpy/f2py/cfuncs.py --- a/numpy/f2py/cfuncs.py +++ b/numpy/f2py/cfuncs.py @@ -312,7 +312,7 @@ needs['pyobj_from_string1']=['string'] cppmacros['pyobj_from_string1']='#define pyobj_from_string1(v) (PyString_FromString((char )v))' needs['pyobj_from_string1size']=['string'] -cppmacros['pyobj_from_string1size']='#define pyobj_from_string1size(v,len) (PyString_FromStringAndSize((char )v, len))' +cppmacros['pyobj_from_string1size']='#define pyobj_from_string1size(v,len) (PyUString_FromStringAndSize((char )v, len))' needs['TRYPYARRAYTEMPLATE']=['PRINTPYOBJERR'] cppmacros['TRYPYARRAYTEMPLATE']="""\ / New SciPy */	f2py sometimes generates python3-incompatible C wrappings f2py uses deprecated CPython functions (notably `PyString_FromStringAndSize`) when wrapping callbacks with string parameters. An example is the FORTRAN-77 subroutine ``` fortran SUBROUTINE BUG(CALLBACK, A) IMPLICIT NONE EXTERNAL CALLBACK DOUBLE PRECISION CALLBACK DOUBLE PRECISION A CHARACTER*1 R R = 'R' !f2py intent(out) A A = CALLBACK(R) END SUBROUTINE ``` compiled with ``` f2py -c -m bug test.f # on Arch Linux, this is the python3 version of f2py ``` with the Python wrapping ``` python #!/usr/bin/env python from __future__ import print_function import bug def callback(code): if code == 'F': return 2.0 else: return 3.0 print(bug.bug(callback)) ``` Calling the Python2 version of `f2py` and running the script under Python 2.7 yields the correct answer. However, under Python3 we get ``` ~/Documents/Code/Fortran/f2pyBug $ python3 test.py Traceback (most recent call last): File "test.py", line 4, in <module> import bug ImportError: /home/drwells/Documents/Code/Fortran/f2pyBug/bug.cpython-33m.so: undefined symbol: PyString_FromStringAndSize ``` I tested this under Arch Linux (fully updated) with Python 3.3.3 and Python 2.7.6. In both cases I used Numpy 1.8. I also tested this under Ubuntu (Quantal) with Python 3.2.3 and Python 2.7.3. In both cases I used Numpy 1.6.2.	Fyi, unfortunately f2py is not regularly maintained at this point, so you may or may not find someone stepping up to fix it. Your best bet might be to see if you can figure out a patch yourself - we'd definitely be interested in merging such a thing. On 3 Feb 2014 13:51, "drwells" notifications@github.com wrote: > f2py uses deprecated CPython functions (notably PyString_FromStringAndSize) > when wrapping callbacks with string parameters. An example is the > FORTRAN-77 subroutine > > ``` > SUBROUTINE BUG(CALLBACK, A) > IMPLICIT NONE EXTERNAL CALLBACK > DOUBLE PRECISION CALLBACK > DOUBLE PRECISION A > CHARACTER1 R > R = 'R'!f2py intent(out) A > A = CALLBACK(R) > END SUBROUTINE > ``` > > compiled with > > f2py -c -m bug test.f # on Arch Linux, this is the python3 version of f2py > > with the Python wrapping > > #!/usr/bin/env pythonfrom future* import print_functionimport bug > def callback(code): > if code == 'F': > return 2.0 > else: > return 3.0 > print(bug.bug(callback)) > > Calling the Python2 version of f2py and running the script under Python > 2.7 yields the correct answer. However, under Python3 we get > > ~/Documents/Code/Fortran/f2pyBug $ python3 test.py > Traceback (most recent call last): > File "test.py", line 4, in <module> > import bug > ImportError: /home/drwells/Documents/Code/Fortran/f2pyBug/bug.cpython-33m.so: undefined symbol: PyString_FromStringAndSize > > I tested this under Arch Linux (fully updated) with Python 3.3.3 and > Python 2.7.6. In both cases I used Numpy 1.8. I also tested this under > Ubuntu (Quantal) with Python 3.2.3 and Python 2.7.3. In both cases I used > Numpy 1.6.2. > > — > Reply to this email directly or view it on GitHubhttps://github.com/numpy/numpy/issues/4256 > . These are likely easy to fix by adding the missing defines to numpy/f2py/src/fortranobject.h	2014-02-16T22:17:50Z	[]	[]	Traceback (most recent call last): File "test.py", line 4, in <module> import bug ImportError: /home/drwells/Documents/Code/Fortran/f2pyBug/bug.cpython-33m.so: undefined symbol: PyString_FromStringAndSize	10,294
numpy/numpy	numpy__numpy-4479	42be275997e56b7e21d3adab5d5a5142876db9f0	diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -1593,6 +1593,7 @@ def __init__(self, pyfunc, otypes='', doc=None, excluded=None, cache=False): self.pyfunc = pyfunc self.cache = cache + self._ufunc = None # Caching to improve default performance if doc is None: self.__doc__ = pyfunc.__doc__ @@ -1616,9 +1617,6 @@ def __init__(self, pyfunc, otypes='', doc=None, excluded=None, excluded = set() self.excluded = set(excluded) - if self.otypes and not self.excluded: - self._ufunc = None # Caching to improve default performance - def __call__(self, args, kwargs): """ Return arrays with the results of `pyfunc` broadcast (vectorized) over @@ -1652,7 +1650,8 @@ def func(vargs): def _get_ufunc_and_otypes(self, func, args): """Return (ufunc, otypes).""" # frompyfunc will fail if args is empty - assert args + if not args: + raise ValueError('args can not be empty') if self.otypes: otypes = self.otypes	_vectorize_call error when otype attribute is set after vectorize object creation if a vectorize object is created without otype argument, an attempt to set it later causes an error: ``` python v = numpy.vectorize( lambda x: x ) v.otypes='i' v( [1,2] ) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1873, in __call__ return self._vectorize_call(func=func, args=vargs) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1933, in _vectorize_call ufunc, otypes = self._get_ufunc_and_otypes(func=func, args=args) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1886, in _get_ufunc_and_otypes if func is self.pyfunc and self._ufunc is not None: AttributeError: 'vectorize' object has no attribute '_ufunc' ```	Still in 1.9-devel. `_get_ufunc_and_otypes` is buggy in other ways, it uses `assert` for flow control.	2014-03-11T08:20:58Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1873, in __call__ return self._vectorize_call(func=func, args=vargs) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1933, in _vectorize_call ufunc, otypes = self._get_ufunc_and_otypes(func=func, args=args) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1886, in _get_ufunc_and_otypes if func is self.pyfunc and self._ufunc is not None: AttributeError: 'vectorize' object has no attribute '_ufunc'	10,310
numpy/numpy	numpy__numpy-4588	a0794f63d548e688e2eed76a9dc4e8df0ea33846	diff --git a/numpy/lib/npyio.py b/numpy/lib/npyio.py --- a/numpy/lib/npyio.py +++ b/numpy/lib/npyio.py @@ -845,6 +845,11 @@ def split_line(line): continue if usecols: vals = [vals[i] for i in usecols] + if len(vals) != N: + line_num = i + skiprows + 1 + raise ValueError("Wrong number of columns at line %d" + % line_num) + # Convert each value according to its column and store items = [conv(val) for (conv, val) in zip(converters, vals)] # Then pack it according to the dtype's nesting	loadtxt should give line numbers of problems (Trac #1998) _Original ticket http://projects.scipy.org/numpy/ticket/1998 on 2011-12-15 by @samtygier, assigned to unknown._ currently input like ``` 1 2 3 4 5 6 78 ``` will cause loadtxt to give the following error ``` Traceback (most recent call last): File "./np_lt.py", line 5, in <module> numpy.loadtxt(sys.argv[1]) File "/usr/lib/python2.7/site-packages/numpy/lib/npyio.py", line 804, in loadtxt X = np.array(X, dtype) ValueError: setting an array element with a sequence. ``` This would be far more useful, if it contained some information about where the problem was. This is similar to the bug #1810 , but for the loadtxt function. I attach a patch that gives the following message: ``` ValueError: Wrong number of columns at line 3 ``` it also provides a test that the input raises the correct error. Inconveniently it is the same error type as before, so the test is not very useful. However I thing ValueError is probably the most appropriate in this case. Maybe there is a way to check the exception message.	Attachment added by @samtygier on 2011-12-15: [numpy_loadtxt.diff](http://projects.scipy.org/numpy/attachment/ticket/1998/numpy_loadtxt.diff) Good idea.	2014-04-05T05:03:53Z	[]	[]	Traceback (most recent call last): File "./np_lt.py", line 5, in <module> numpy.loadtxt(sys.argv[1]) File "/usr/lib/python2.7/site-packages/numpy/lib/npyio.py", line 804, in loadtxt X = np.array(X, dtype) ValueError: setting an array element with a sequence.	10,314
numpy/numpy	numpy__numpy-4677	8e0ac440329188b959520a0a7ce41ef60b2fb3c2	diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -7018,8 +7018,6 @@ def asarray(a, dtype=None, order=None): <class 'numpy.ma.core.MaskedArray'> """ - if dtype is None and type(a) is MaskedArray: - return a return masked_array(a, dtype=dtype, copy=False, keep_mask=True, subok=False) def asanyarray(a, dtype=None): @@ -7065,8 +7063,6 @@ def asanyarray(a, dtype=None): <class 'numpy.ma.core.MaskedArray'> """ - if dtype is None and isinstance(a, MaskedArray): - return a return masked_array(a, dtype=dtype, copy=False, keep_mask=True, subok=True)	Test error for scipy 0.15.0.dev-5d197ed The first bad commit is d8fd28389adb491e24b7cdc25cd1b20f539310c3, the isolated error is ``` Traceback (most recent call last): File "/home/charris/scipy-test-fail.py", line 53, in test_trim assert_equal(trimx._mask.ravel(), expected) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 123, in assert_equal return assert_array_equal(actual, desired, err_msg) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 196, in assert_array_equal header='Arrays are not equal') File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 189, in assert_array_compare verbose=verbose, header=header) File "/home/charris/.local/lib/python2.7/site-packages/numpy/testing/utils.py", line 651, in assert_array_compare raise AssertionError(msg) AssertionError: Arrays are not equal (mismatch 9.09090909091%) x: array([ True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, False, False, False, False, False, False, False,... y: array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,... ---------------------------------------------------------------------- Ran 1 test in 0.005s ```	Looks like the problem is returning a view rather than a copy. Or more precisely, not making a copy and returning the original array. Probably the easiest fix is to revert this change, at least for numpy 1.9.	2014-05-05T17:28:43Z	[]	[]	Traceback (most recent call last): File "/home/charris/scipy-test-fail.py", line 53, in test_trim assert_equal(trimx._mask.ravel(), expected) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 123, in assert_equal return assert_array_equal(actual, desired, err_msg) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 196, in assert_array_equal header='Arrays are not equal') File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 189, in assert_array_compare verbose=verbose, header=header) File "/home/charris/.local/lib/python2.7/site-packages/numpy/testing/utils.py", line 651, in assert_array_compare raise AssertionError(msg) AssertionError:	10,318
numpy/numpy	numpy__numpy-4792	db710cefeecf51d6253e421712726c1506a6f65b	diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -651,7 +651,7 @@ def piecewise(x, condlist, funclist, args, kw): The output is the same shape and type as x and is found by calling the functions in `funclist` on the appropriate portions of `x`, as defined by the boolean arrays in `condlist`. Portions not covered - by any condition have undefined values. + by any condition have a default value of 0. See Also @@ -693,32 +693,24 @@ def piecewise(x, condlist, funclist, args, **kw): if (isscalar(condlist) or not (isinstance(condlist[0], list) or isinstance(condlist[0], ndarray))): condlist = [condlist] - condlist = [asarray(c, dtype=bool) for c in condlist] + condlist = array(condlist, dtype=bool) n = len(condlist) - if n == n2 - 1: # compute the "otherwise" condition. - totlist = condlist[0] - for k in range(1, n): - totlist \|= condlist[k] - condlist.append(~totlist) - n += 1 - if (n != n2): - raise ValueError( - "function list and condition list must be the same") - zerod = False # This is a hack to work around problems with NumPy's # handling of 0-d arrays and boolean indexing with # numpy.bool_ scalars + zerod = False if x.ndim == 0: x = x[None] zerod = True - newcondlist = [] - for k in range(n): - if condlist[k].ndim == 0: - condition = condlist[k][None] - else: - condition = condlist[k] - newcondlist.append(condition) - condlist = newcondlist + if condlist.shape[-1] != 1: + condlist = condlist.T + if n == n2 - 1: # compute the "otherwise" condition. + totlist = np.logical_or.reduce(condlist, axis=0) + condlist = np.vstack([condlist, ~totlist]) + n += 1 + if (n != n2): + raise ValueError( + "function list and condition list must be the same") y = zeros(x.shape, x.dtype) for k in range(n):	Fixed bug in numpy.piecewise() for 0-d array handling Updated numpy/lib/function_base.py to fix bug in numpy.piecewise() for 0-d array handling and boolean indexing with scalars. Bug test case: ``` >>> numpy.piecewise(5, [True, False], [1, 0]) Traceback (most recent call last): [...] y[condlist[k]] = item ValueError: boolean index array has too many values ``` After fix: ``` >>> numpy.piecewise(5, [True, False], [1, 0]) array(1) ```	This pull request [fails](http://travis-ci.org/numpy/numpy/builds/1788960) (merged e553e1b4 into 731cf3aa). The current version of this pull request appears to fail the existing tests (e.g., this build reports failures: http://travis-ci.org/#!/numpy/numpy/jobs/1788964), so that would need to be fixed. Also, some sort of test needs to be added to make sure that the bug is fixed (and stays that way). Hi njsmith, thanks. I've fixed both my change to numpy.piecewise() as well as the test cases to cover both my bug case above as well as making sure that the tests are all logical. This pull request [passes](http://travis-ci.org/numpy/numpy/builds/1804356) (merged e2ad1a72 into 731cf3aa). I guess you rearranged the order of all the tests in the source? It makes the diff unreadable unfortunately -- I can't tell at all what you've actually changed without reverse-engineering it by hand, and I'm afraid I'm too lazy for that. Can you revert the no-op changes, or split them into a separate commit, or at least describe what changes were actually made? Is this issue (i.e. this pull request) going to be fixed anytime soon? It's a bit anoying to construct a 1d array with one element each time I want to call this function with a scalar. I guess it will be fixed as soon as someone does the work :-) Hm, got it :) Should an issue be filed first? The main point of filing an issue is to make sure that something doesn't get forgotten, and this PR is already sort of doing that (i.e., it shows up on the list of open issues if anyone goes to look at it). But feel free to file one if you like. If you're thinking about fixing the bug yourself, then we're pretty laid back about these things -- you can just go ahead and submit a new PR without filing an issue if you want. We'd rather get the fix than worry about the process :-). The main thing that needs done is just to rewrite the unit tests. In my own testing and usage (and when I've tested with the original unit tests), the code works fine, but I just haven't had the time to sit down and rewrite them to be more similar to the original set and still be complete and systematic. I may get to it next week sometime, but until then I'm swamped. -- Eric On Nov 15, 2012, at 9:33 AM, njsmith notifications@github.com wrote: > The main point of filing an issue is to make sure that something doesn't get forgotten, and this PR is already sort of doing that (i.e., it shows up on the list of open issues if anyone goes to look at it). But feel free to file one if you like. If you're thinking about fixing the bug yourself, then we're pretty laid back about these things -- you can just go ahead and submit a new PR without filing an issue if you want. We'd rather get the fix than worry about the process :-). > > — > Reply to this email directly or view it on GitHub. That's one of the things I was about to ask, it is ok to rewrite the tests? Because right now IMHO they don't seem very meaningful for the purpose of the function. I'd like to try it myself this weekend, or next one as much. It's OK to make any changes whatsoever, the requirements are just that you should have a reason, and we should be able to look at your changes and figure out if they're good :-). How exactly to do that depends on the change. Easily readable diffs, English explanations of what you did and why... whatever works. On 17 Nov 2012 00:02, "Juan Luis Cano Rodríguez" notifications@github.com wrote: > That's one of the things I was about to ask, it is ok to rewrite the > tests? Because right now IMHO they don't seem very meaningful for the > purpose of the function. I'd like to try it myself this weekend, or next > one as much. > > — > Reply to this email directly or view it on GitHubhttps://github.com/numpy/numpy/pull/331#issuecomment-10467386. What is the status of this? Well actually I didn't do it any of the weekends as you can notice :( But anyway now that you ask about it, if I recall correctly there was some discussion recently about rank-0 arrays in the mailing list but I am not able to find it. Are there any constraints on how to do this? Other than that it would be my first contribution to NumPy and I am a bit unsecure but I am willing to give it a try. Yes, I was referring to this: http://www.mail-archive.com/numpy-discussion@scipy.org/msg39968.html My question was if there are any near future plans for rank-0 arrays that relate this and other issues in such a way that it requires changing low-level code or if we can address this individually. No, I don't think there will be any near-future changes in how 0d arrays work that will affect this bug. On 1 Mar 2013 07:29, "Juan Luis Cano Rodríguez" notifications@github.com wrote: > Yes, I was referring to this: > > http://www.mail-archive.com/numpy-discussion@scipy.org/msg39968.html > > My question was if there are any near future plans for rank-0 arrays that > relate this and other issues in such a way that it requires changing > low-level code or if we can address this individually. > > — > Reply to this email directly or view it on GitHubhttps://github.com/numpy/numpy/pull/331#issuecomment-14276274 > . Sorry for the delay of months. Finishing up my degree and job hunting, so I've been...occupied. I reverted most of the unit test changes so that it's clearer what I changed. Maybe I'll re-up the unit test changes at some point, but this particular commit should be easier to understand when you diff between this state and the state 3 commits ago. @njsmith Can this go in. Ping Travisbot just because it was a long time ago. @ericsuh Made some comments. The commits should also be squashed into one. I've put up a cleaned version at https://github.com/charris/numpy in the branch gh-331 Needs finishing up. I seem to have deleted my cleanup somewhere along the line. In the docstring: "Each boolean array in `condlist` selects a piece of `x`, and should therefore be of the same shape as `x`." But actually even the tests do not respect this (and still everything works): https://github.com/numpy/numpy/blob/b1c69df01b673cc086065112da6780d8548a0dfa/numpy/lib/tests/test_function_base.py#L1468 There are other bugs too. For instance, there's `test_0d` but even though `np.piecewise(x, x > 3, [4, 0])` passes, `np.piecewise(x, [x > 3, x <= 3], [4, 0])` fails (same error as originally reported). I don't think this behaviour is consistent - that's what I actually meant a year and a half ago. The precondition stated in the docstring makes sense for me and should be somehow enforced - is anybody against it? I cannot promise anything but I may try and solve this at once.	2014-06-07T20:44:26Z	[]	[]	Traceback (most recent call last): [...] y[condlist[k]] = item ValueError: boolean index array has too many values	10,321
numpy/numpy	numpy__numpy-5149	58350f4608a22f4b4b66795f51eaefc206bd02b8	diff --git a/numpy/ma/extras.py b/numpy/ma/extras.py --- a/numpy/ma/extras.py +++ b/numpy/ma/extras.py @@ -434,8 +434,10 @@ def apply_over_axes(func, a, axes): raise ValueError("function is not returning " "an array of the correct shape") return val -apply_over_axes.__doc__ = np.apply_over_axes.__doc__[ - :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ + +if apply_over_axes.__doc__ is not None: + apply_over_axes.__doc__ = np.apply_over_axes.__doc__[ + :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ """ Examples @@ -462,7 +464,7 @@ def apply_over_axes(func, a, axes): [[[46] [--] [124]]] -""" + """ def average(a, axis=None, weights=None, returned=False):	Numpy crashes with -OO Hello, I ran `$ python -OO -c 'import numpy'` on a fresh miniconda install (after `$ conda install numpy`) and I got the following output: ``` Traceback (most recent call last): File "<string>", line 1, in <module> File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/__init__.py", line 191, in <module> from . import ma File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/__init__.py", line 49, in <module> from . import extras File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/extras.py", line 438, in <module> :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ AttributeError: 'NoneType' object has no attribute 'find' ``` Running `$ python` gives me the following output (I am running OS X Mavericks on a Dec-2010 MacBook Pro): ``` Python 2.7.8 \|Continuum Analytics, Inc.\| (default, Aug 21 2014, 15:21:46) [GCC 4.2.1 (Apple Inc. build 5577)] on darwin Type "help", "copyright", "credits" or "license" for more information. Anaconda is brought to you by Continuum Analytics. Please check out: http://continuum.io/thanks and https://binstar.org ``` Finally, running `python -c 'import numpy'` works normally. I hope this is enough information for you all, but please let me know if you need more. (It's the second bug report I submit in my life, and the first one was about a typo in documentation!) Thanks a lot for making numpy available to us, it is a very, very useful tool that works greatly for me.		2014-10-04T02:12:20Z	[]	[]	Traceback (most recent call last): File "<string>", line 1, in <module> File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/__init__.py", line 191, in <module> from . import ma File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/__init__.py", line 49, in <module> from . import extras File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/extras.py", line 438, in <module> :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ AttributeError: 'NoneType' object has no attribute 'find'	10,330
numpy/numpy	numpy__numpy-5455	14445500bdf67600f926c6426bad55977441dca0	diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -145,10 +145,15 @@ class MaskError(MAError): 'S' : 'N/A', 'u' : 999999, 'V' : '???', - 'U' : 'N/A', - 'M8[D]' : np.datetime64('NaT', 'D'), - 'M8[us]' : np.datetime64('NaT', 'us') + 'U' : 'N/A' } + +# Add datetime64 and timedelta64 types +for v in ["Y", "M", "W", "D", "h", "m", "s", "ms", "us", "ns", "ps", + "fs", "as"]: + default_filler["M8[" + v + "]"] = np.datetime64("NaT", v) + default_filler["m8[" + v + "]"] = np.timedelta64("NaT", v) + max_filler = ntypes._minvals max_filler.update([(k, -np.inf) for k in [np.float32, np.float64]]) min_filler = ntypes._maxvals @@ -194,7 +199,7 @@ def default_fill_value(obj): 999999 >>> np.ma.default_fill_value(np.array([1.1, 2., np.pi])) 1e+20 - >>> np.ma.default_fill_value(np.dtype(complex)) + >>> np.ma.default_fill_value(np.dtype(complex)) (1e+20+0j) """ @@ -203,7 +208,7 @@ def default_fill_value(obj): elif isinstance(obj, np.dtype): if obj.subdtype: defval = default_filler.get(obj.subdtype[0].kind, '?') - elif obj.kind == 'M': + elif obj.kind in 'Mm': defval = default_filler.get(obj.str[1:], '?') else: defval = default_filler.get(obj.kind, '?')	Masked array view fails if structured dtype has datetime component A view as `numpy.ma.MaskedArray` fails if the array has a structured dtype, including at least one part that is `datetime64`, as follows: ``` $ python3.3 Python 3.3.2+ (default, Feb 28 2014, 00:52:16) [GCC 4.8.1] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import numpy >>> import numpy, numpy.ma, numpy.version >>> numpy.version.version '1.8.0' >>> A = numpy.empty(shape=(5,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")]) >>> A.view(numpy.ma.MaskedArray) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0 ```	A simpler `numpy.ma.MaskedArray([], shape=(0,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")])` fails with the same `ValueError`. Just tested with with the latest git repository. Problem still exists.	2015-01-15T00:07:57Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0	10,346
numpy/numpy	numpy__numpy-5498	54ebec335384ec1d6f8e65bdd35c4f359797dc0b	diff --git a/numpy/add_newdocs.py b/numpy/add_newdocs.py --- a/numpy/add_newdocs.py +++ b/numpy/add_newdocs.py @@ -885,7 +885,7 @@ def luf(lamdaexpr, args, *kwargs): >>> np.zeros(5) array([ 0., 0., 0., 0., 0.]) - >>> np.zeros((5,), dtype=numpy.int) + >>> np.zeros((5,), dtype=np.int) array([0, 0, 0, 0, 0]) >>> np.zeros((2, 1))	Error in help documentation of numpy.zeros The help documentation for numpy zeros method (`import numpy; help(numpy.zeros)`) shows an example like this: ``` python >>> np.zeros((5,), dtype=numpy.int) array([0, 0, 0, 0, 0]) ``` The documentation throughout `multiarray` module assumes that `numpy` is imported as `np` but the value for `dtype` keyword is passed as `numpy.int`. While the above snippet conveys the point across, it will throw a name error if some one actually tried to execute it: ``` python >>> import numpy as np >>> np.zeros((5,), dtype=numpy.int) Traceback (most recent call last): File "<stdin>", line 1, in <module> NameError: name 'numpy' is not defined ``` This can be fixed by consitent usage of `numpy` or `np`.		2015-01-24T09:50:00Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> NameError: name 'numpy' is not defined	10,351
numpy/numpy	numpy__numpy-5500	1cac77b27d10217aaeb468abb71ed49f241ce469	diff --git a/numpy/lib/npyio.py b/numpy/lib/npyio.py --- a/numpy/lib/npyio.py +++ b/numpy/lib/npyio.py @@ -717,7 +717,8 @@ def loadtxt(fname, dtype=float, comments='#', delimiter=None, """ # Type conversions for Py3 convenience - comments = asbytes(comments) + if comments is not None: + comments = asbytes(comments) user_converters = converters if delimiter is not None: delimiter = asbytes(delimiter) @@ -790,7 +791,10 @@ def pack_items(items, packing): def split_line(line): """Chop off comments, strip, and split at delimiter.""" - line = asbytes(line).split(comments)[0].strip(asbytes('\r\n')) + if comments is None: + line = asbytes(line).strip(asbytes('\r\n')) + else: + line = asbytes(line).split(comments)[0].strip(asbytes('\r\n')) if line: return line.split(delimiter) else: diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -145,10 +145,15 @@ class MaskError(MAError): 'S' : 'N/A', 'u' : 999999, 'V' : '???', - 'U' : 'N/A', - 'M8[D]' : np.datetime64('NaT', 'D'), - 'M8[us]' : np.datetime64('NaT', 'us') + 'U' : 'N/A' } + +# Add datetime64 and timedelta64 types +for v in ["Y", "M", "W", "D", "h", "m", "s", "ms", "us", "ns", "ps", + "fs", "as"]: + default_filler["M8[" + v + "]"] = np.datetime64("NaT", v) + default_filler["m8[" + v + "]"] = np.timedelta64("NaT", v) + max_filler = ntypes._minvals max_filler.update([(k, -np.inf) for k in [np.float32, np.float64]]) min_filler = ntypes._maxvals @@ -194,7 +199,7 @@ def default_fill_value(obj): 999999 >>> np.ma.default_fill_value(np.array([1.1, 2., np.pi])) 1e+20 - >>> np.ma.default_fill_value(np.dtype(complex)) + >>> np.ma.default_fill_value(np.dtype(complex)) (1e+20+0j) """ @@ -203,7 +208,7 @@ def default_fill_value(obj): elif isinstance(obj, np.dtype): if obj.subdtype: defval = default_filler.get(obj.subdtype[0].kind, '?') - elif obj.kind == 'M': + elif obj.kind in 'Mm': defval = default_filler.get(obj.str[1:], '?') else: defval = default_filler.get(obj.kind, '?')	loadtxt(comments=None) considers the string 'None' as a comment symbol Numpy 1.8.2 `numpy.loadtxt(comments=None)` considers the string `'None'` as a comment symbol. Expected behaviour: `comments=None` should indicate that there are no comments symbols. Certainly not cast `None` to a string, what is this, Javascript? Masked array view fails if structured dtype has datetime component A view as `numpy.ma.MaskedArray` fails if the array has a structured dtype, including at least one part that is `datetime64`, as follows: ``` $ python3.3 Python 3.3.2+ (default, Feb 28 2014, 00:52:16) [GCC 4.8.1] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import numpy >>> import numpy, numpy.ma, numpy.version >>> numpy.version.version '1.8.0' >>> A = numpy.empty(shape=(5,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")]) >>> A.view(numpy.ma.MaskedArray) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0 ```	A simpler `numpy.ma.MaskedArray([], shape=(0,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")])` fails with the same `ValueError`. Just tested with with the latest git repository. Problem still exists.	2015-01-25T16:09:27Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0	10,352
numpy/numpy	numpy__numpy-5548	30d755d8737505717d54ed32501261bb94130a7f	diff --git a/numpy/core/_internal.py b/numpy/core/_internal.py --- a/numpy/core/_internal.py +++ b/numpy/core/_internal.py @@ -305,6 +305,174 @@ def _index_fields(ary, fields): copy_dtype = {'names':view_dtype['names'], 'formats':view_dtype['formats']} return array(view, dtype=copy_dtype, copy=True) +def _get_all_field_offsets(dtype, base_offset=0): + """ Returns the types and offsets of all fields in a (possibly structured) + data type, including nested fields and subarrays. + + Parameters + ---------- + dtype : data-type + Data type to extract fields from. + base_offset : int, optional + Additional offset to add to all field offsets. + + Returns + ------- + fields : list of (data-type, int) pairs + A flat list of (dtype, byte offset) pairs. + + """ + fields = [] + if dtype.fields is not None: + for name in dtype.names: + sub_dtype = dtype.fields[name][0] + sub_offset = dtype.fields[name][1] + base_offset + fields.extend(_get_all_field_offsets(sub_dtype, sub_offset)) + else: + if dtype.shape: + sub_offsets = _get_all_field_offsets(dtype.base, base_offset) + count = 1 + for dim in dtype.shape: + count = dim + fields.extend((typ, off + dtype.base.itemsizej) + for j in range(count) for (typ, off) in sub_offsets) + else: + fields.append((dtype, base_offset)) + return fields + +def _check_field_overlap(new_fields, old_fields): + """ Perform object memory overlap tests for two data-types (see + _view_is_safe). + + This function checks that new fields only access memory contained in old + fields, and that non-object fields are not interpreted as objects and vice + versa. + + Parameters + ---------- + new_fields : list of (data-type, int) pairs + Flat list of (dtype, byte offset) pairs for the new data type, as + returned by _get_all_field_offsets. + old_fields: list of (data-type, int) pairs + Flat list of (dtype, byte offset) pairs for the old data type, as + returned by _get_all_field_offsets. + + Raises + ------ + TypeError + If the new fields are incompatible with the old fields + + """ + from .numerictypes import object_ + from .multiarray import dtype + + #first go byte by byte and check we do not access bytes not in old_fields + new_bytes = set() + for tp, off in new_fields: + new_bytes.update(set(range(off, off+tp.itemsize))) + old_bytes = set() + for tp, off in old_fields: + old_bytes.update(set(range(off, off+tp.itemsize))) + if new_bytes.difference(old_bytes): + raise TypeError("view would access data parent array doesn't own") + + #next check that we do not interpret non-Objects as Objects, and vv + obj_offsets = [off for (tp, off) in old_fields if tp.type is object_] + obj_size = dtype(object_).itemsize + + for fld_dtype, fld_offset in new_fields: + if fld_dtype.type is object_: + # check we do not create object views where + # there are no objects. + if fld_offset not in obj_offsets: + raise TypeError("cannot view non-Object data as Object type") + else: + # next check we do not create non-object views + # where there are already objects. + # see validate_object_field_overlap for a similar computation. + for obj_offset in obj_offsets: + if (fld_offset < obj_offset + obj_size and + obj_offset < fld_offset + fld_dtype.itemsize): + raise TypeError("cannot view Object as non-Object type") + +def _getfield_is_safe(oldtype, newtype, offset): + """ Checks safety of getfield for object arrays. + + As in _view_is_safe, we need to check that memory containing objects is not + reinterpreted as a non-object datatype and vice versa. + + Parameters + ---------- + oldtype : data-type + Data type of the original ndarray. + newtype : data-type + Data type of the field being accessed by ndarray.getfield + offset : int + Offset of the field being accessed by ndarray.getfield + + Raises + ------ + TypeError + If the field access is invalid + + """ + new_fields = _get_all_field_offsets(newtype, offset) + old_fields = _get_all_field_offsets(oldtype) + # raises if there is a problem + _check_field_overlap(new_fields, old_fields) + +def _view_is_safe(oldtype, newtype): + """ Checks safety of a view involving object arrays, for example when + doing:: + + np.zeros(10, dtype=oldtype).view(newtype) + + We need to check that + 1) No memory that is not an object will be interpreted as a object, + 2) No memory containing an object will be interpreted as an arbitrary type. + Both cases can cause segfaults, eg in the case the view is written to. + Strategy here is to also disallow views where newtype has any field in a + place oldtype doesn't. + + Parameters + ---------- + oldtype : data-type + Data type of original ndarray + newtype : data-type + Data type of the view + + Raises + ------ + TypeError + If the new type is incompatible with the old type. + + """ + new_fields = _get_all_field_offsets(newtype) + new_size = newtype.itemsize + + old_fields = _get_all_field_offsets(oldtype) + old_size = oldtype.itemsize + + # if the itemsizes are not equal, we need to check that all the + # 'tiled positions' of the object match up. Here, we allow + # for arbirary itemsizes (even those possibly disallowed + # due to stride/data length issues). + if old_size == new_size: + new_num = old_num = 1 + else: + gcd_new_old = _gcd(new_size, old_size) + new_num = old_size // gcd_new_old + old_num = new_size // gcd_new_old + + # get position of fields within the tiling + new_fieldtile = [(tp, off + new_sizej) + for j in range(new_num) for (tp, off) in new_fields] + old_fieldtile = [(tp, off + old_sizej) + for j in range(old_num) for (tp, off) in old_fields] + + # raises if there is a problem + _check_field_overlap(new_fieldtile, old_fieldtile) + # Given a string containing a PEP 3118 format specifier, # construct a Numpy dtype	TypeError: Cannot change data-type for object array If I try and read in the array contained in this npy file: https://gist.github.com/astrofrog/8c2d188005f31e0bba36/raw/3065c8fa220a6eaccbff20565d0d520c07e5e7e6/test.npy then try and print out the array, so: ``` python import numpy as np array = np.load('test.npy') print(array) ``` I get: ``` Traceback (most recent call last): File "test2.py", line 5, in <module> print(array) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/numeric.py", line 1767, in array_str return array2string(a, max_line_width, precision, suppress_small, ' ', "", str) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 459, in array2string separator, prefix, formatter=formatter) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 329, in _array2string _summaryEdgeItems, summary_insert)[:-1] File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 526, in _formatArray s += _formatArray(a[-i], format_function, rank-1, max_line_len, File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 481, in __getitem__ return obj.view(dtype=(self.dtype.type, obj.dtype.descr)) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 540, in view return ndarray.view(self, dtype) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 457, in __setattr__ raise exctype(value) TypeError: Cannot change data-type for object array. ``` This is with the latest developer version of Numpy (3c5409e4e38e6034d69d0042bf2a3bc854ef2e53) and Python 3.4 on MacOS X. The dtype can be printed, as can individual columns, but the array as a whole can't. This doesn't occur in the latest stable release so may be a regression?		2015-02-09T05:49:41Z	[]	[]	Traceback (most recent call last): File "test2.py", line 5, in <module> print(array) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/numeric.py", line 1767, in array_str return array2string(a, max_line_width, precision, suppress_small, ' ', "", str) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 459, in array2string separator, prefix, formatter=formatter) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 329, in _array2string _summaryEdgeItems, summary_insert)[:-1] File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 526, in _formatArray s += _formatArray(a[-i], format_function, rank-1, max_line_len, File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 481, in __getitem__ return obj.view(dtype=(self.dtype.type, obj.dtype.descr)) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 540, in view return ndarray.view(self, dtype) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 457, in __setattr__ raise exctype(value) TypeError: Cannot change data-type for object array.	10,357
numpy/numpy	numpy__numpy-5584	2e016ac65aceab4e08217794d6be7b365793976a	diff --git a/numpy/core/fromnumeric.py b/numpy/core/fromnumeric.py --- a/numpy/core/fromnumeric.py +++ b/numpy/core/fromnumeric.py @@ -691,8 +691,16 @@ def argpartition(a, kth, axis=-1, kind='introselect', order=None): >>> x[np.argpartition(x, (1, 3))] array([1, 2, 3, 4]) + >>> x = [3, 4, 2, 1] + >>> np.array(x)[np.argpartition(x, 3)] + array([2, 1, 3, 4]) + """ - return a.argpartition(kth, axis, kind=kind, order=order) + try: + argpartition = a.argpartition + except AttributeError: + return _wrapit(a, 'argpartition',kth, axis, kind, order) + return argpartition(kth, axis, kind=kind, order=order) def sort(a, axis=-1, kind='quicksort', order=None):	argpartition fails on non-ndarray array-likes While `partition` works, as advertised, array-likeinputs like, e.g. lists: ``` >>> np.partition([5, 4, 3, 2, 1], 2) array([1, 2, 3, 4, 5]) ``` `argpartition` raises an error if its first argument is not an `ndarray`: ``` >>> np.argpartition([5, 4, 3, 2, 1], 2) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition' ```	If any array-like inputs are to be accepted, then as in `argsort`, this can be implemented: ``` try: argpartition = a.argpartition except AttributeError: return _wrapit(a, 'argpartition', kth, axis, kind, order) return argpartition(kth, axis, kind=kind, order=order) ``` else it can be like `sort` and `partition`: ``` if axis is None: a = asanyarray(a).flatten() axis = 0 else: a = asanyarray(a).copy(order="K") return a.argpartition(kth, axis, kind=kind, order=order) ``` Just asking out of curiosity, the following example is shown in the documentation : ``` >>> x = np.array([3, 4, 2, 1]) >>> x[np.argpartition(x, 3)] array([2, 1, 3, 4]) ``` Now, if any array-like inputs are allowed, then the following behavior is needed. ``` >>> x = [3, 4, 2, 1] >>>np.array(x)[np.argpartition(x, 3)] array([2, 1, 3, 4]) >>>x[np.argpartition(x, 3)] Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: only integer arrays with one element can be converted to an index ``` If this is the procedure to be followed, as in case of `argsort`, I think it should be documented. Please correct me if no such use is needed in general scenarios. it probably should be like argsort, Don't remember why its not, maybe just a mistake, I can't remember there being a reason. compared to sort, partition can't be overridden from the C side but that shouldn't matter here. I think the `partition` and `sort` C API functions are defined in `item_selection.c`. Can you please clarify on what `compared to sort, partition can't be overridden from the C side` means ? Could I have a go at fixing this ? no, that is kind of intentional the C sort functions are part of a public struct, adding partition would break the ABI and just add more stuff that should not be in that place. So I went with a internal lookupt able that could get hooks for user functions. But I don't think its really worthwhile, I don't think anybody would use it. @juliantaylor Thanks for the clarification. I am new to working with these kind of projects and wasn't aware of the nature of public struct functions and the way in which they are maintained.	2015-02-19T13:56:08Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition'	10,359
numpy/numpy	numpy__numpy-5616	bf6f80d994a154f25eb5e8beca5babdf31b13eca	diff --git a/numpy/core/fromnumeric.py b/numpy/core/fromnumeric.py --- a/numpy/core/fromnumeric.py +++ b/numpy/core/fromnumeric.py @@ -679,8 +679,16 @@ def argpartition(a, kth, axis=-1, kind='introselect', order=None): >>> x[np.argpartition(x, (1, 3))] array([1, 2, 3, 4]) + >>> x = [3, 4, 2, 1] + >>> np.array(x)[np.argpartition(x, 3)] + array([2, 1, 3, 4]) + """ - return a.argpartition(kth, axis, kind=kind, order=order) + try: + argpartition = a.argpartition + except AttributeError: + return _wrapit(a, 'argpartition',kth, axis, kind, order) + return argpartition(kth, axis, kind=kind, order=order) def sort(a, axis=-1, kind='quicksort', order=None):	argpartition fails on non-ndarray array-likes While `partition` works, as advertised, array-likeinputs like, e.g. lists: ``` >>> np.partition([5, 4, 3, 2, 1], 2) array([1, 2, 3, 4, 5]) ``` `argpartition` raises an error if its first argument is not an `ndarray`: ``` >>> np.argpartition([5, 4, 3, 2, 1], 2) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition' ```	If any array-like inputs are to be accepted, then as in `argsort`, this can be implemented: ``` try: argpartition = a.argpartition except AttributeError: return _wrapit(a, 'argpartition', kth, axis, kind, order) return argpartition(kth, axis, kind=kind, order=order) ``` else it can be like `sort` and `partition`: ``` if axis is None: a = asanyarray(a).flatten() axis = 0 else: a = asanyarray(a).copy(order="K") return a.argpartition(kth, axis, kind=kind, order=order) ``` Just asking out of curiosity, the following example is shown in the documentation : ``` >>> x = np.array([3, 4, 2, 1]) >>> x[np.argpartition(x, 3)] array([2, 1, 3, 4]) ``` Now, if any array-like inputs are allowed, then the following behavior is needed. ``` >>> x = [3, 4, 2, 1] >>>np.array(x)[np.argpartition(x, 3)] array([2, 1, 3, 4]) >>>x[np.argpartition(x, 3)] Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: only integer arrays with one element can be converted to an index ``` If this is the procedure to be followed, as in case of `argsort`, I think it should be documented. Please correct me if no such use is needed in general scenarios. it probably should be like argsort, Don't remember why its not, maybe just a mistake, I can't remember there being a reason. compared to sort, partition can't be overridden from the C side but that shouldn't matter here. I think the `partition` and `sort` C API functions are defined in `item_selection.c`. Can you please clarify on what `compared to sort, partition can't be overridden from the C side` means ? Could I have a go at fixing this ? no, that is kind of intentional the C sort functions are part of a public struct, adding partition would break the ABI and just add more stuff that should not be in that place. So I went with a internal lookupt able that could get hooks for user functions. But I don't think its really worthwhile, I don't think anybody would use it. @juliantaylor Thanks for the clarification. I am new to working with these kind of projects and wasn't aware of the nature of public struct functions and the way in which they are maintained.	2015-02-28T13:04:47Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition'	10,361
numpy/numpy	numpy__numpy-5742	147c60f83f401037ff29593826d2c5729a73c2c5	diff --git a/numpy/lib/type_check.py b/numpy/lib/type_check.py --- a/numpy/lib/type_check.py +++ b/numpy/lib/type_check.py @@ -324,12 +324,13 @@ def nan_to_num(x): Returns ------- - out : ndarray, float - Array with the same shape as `x` and dtype of the element in `x` with - the greatest precision. NaN is replaced by zero, and infinity - (-infinity) is replaced by the largest (smallest or most negative) - floating point value that fits in the output dtype. All finite numbers - are upcast to the output dtype (default float64). + out : ndarray + New Array with the same shape as `x` and dtype of the element in + `x` with the greatest precision. If `x` is inexact, then NaN is + replaced by zero, and infinity (-infinity) is replaced by the + largest (smallest or most negative) floating point value that fits + in the output dtype. If `x` is not inexact, then a copy of `x` is + returned. See Also -------- @@ -354,33 +355,22 @@ def nan_to_num(x): -1.28000000e+002, 1.28000000e+002]) """ - try: - t = x.dtype.type - except AttributeError: - t = obj2sctype(type(x)) - if issubclass(t, _nx.complexfloating): - return nan_to_num(x.real) + 1j * nan_to_num(x.imag) - else: - try: - y = x.copy() - except AttributeError: - y = array(x) - if not issubclass(t, _nx.integer): - if not y.shape: - y = array([x]) - scalar = True - else: - scalar = False - are_inf = isposinf(y) - are_neg_inf = isneginf(y) - are_nan = isnan(y) - maxf, minf = _getmaxmin(y.dtype.type) - y[are_nan] = 0 - y[are_inf] = maxf - y[are_neg_inf] = minf - if scalar: - y = y[0] - return y + x = _nx.array(x, subok=True) + xtype = x.dtype.type + if not issubclass(xtype, _nx.inexact): + return x + + iscomplex = issubclass(xtype, _nx.complexfloating) + isscalar = (x.ndim == 0) + + x = x[None] if isscalar else x + dest = (x.real, x.imag) if iscomplex else (x,) + maxf, minf = _getmaxmin(x.real.dtype) + for d in dest: + _nx.copyto(d, 0.0, where=isnan(d)) + _nx.copyto(d, maxf, where=isposinf(d)) + _nx.copyto(d, minf, where=isneginf(d)) + return x[0] if isscalar else x #-----------------------------------------------------------------------------	An error is given if a integer list is passed to nan_to_num (Trac #880) _Original ticket http://projects.scipy.org/numpy/ticket/880 on 2008-08-06 by @bsouthey, assigned to unknown._ When an integer list is passed to nan_to_num then an error occurs. This does not occur if the list contains floats or a numpy integer array is used. ``` np.nan_to_num([1.0,3]) # returns: array([ 1., 3.]) n=np.array([1,3]) np.nan_to_num(n) # returns: array([1, 3]) np.nan_to_num([1,3]) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 135, in nan_to_num maxf, minf = _getmaxmin(y.dtype.type) File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 103, in _getmaxmin f = getlimits.finfo(t) File "/usr/lib64/python2.5/site-packages/numpy/lib/getlimits.py", line 46, in __new__ raise ValueError, "data type %r not inexact" % (dtype) ValueError: data type <type 'numpy.int64'> not inexact ```	Milestone changed to `1.4.0` by @cournape on 2009-03-09 Milestone changed to `Unscheduled` by @mwiebe on 2011-03-23 Still present in 1.9-devel.	2015-04-03T17:34:50Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 135, in nan_to_num maxf, minf = _getmaxmin(y.dtype.type) File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 103, in _getmaxmin f = getlimits.finfo(t) File "/usr/lib64/python2.5/site-packages/numpy/lib/getlimits.py", line 46, in __new__ raise ValueError, "data type %r not inexact" % (dtype) ValueError: data type <type 'numpy.int64'> not inexact	10,367
numpy/numpy	numpy__numpy-6556	2d899ea2301e155b06acf585020866ca1953bce5	diff --git a/numpy/lib/stride_tricks.py b/numpy/lib/stride_tricks.py --- a/numpy/lib/stride_tricks.py +++ b/numpy/lib/stride_tricks.py @@ -62,11 +62,14 @@ def _broadcast_to(array, shape, subok, readonly): if any(size < 0 for size in shape): raise ValueError('all elements of broadcast shape must be non-' 'negative') + needs_writeable = not readonly and array.flags.writeable + extras = ['reduce_ok'] if needs_writeable else [] + op_flag = 'readwrite' if needs_writeable else 'readonly' broadcast = np.nditer( - (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'], - op_flags=['readonly'], itershape=shape, order='C').itviews[0] + (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'] + extras, + op_flags=[op_flag], itershape=shape, order='C').itviews[0] result = _maybe_view_as_subclass(array, broadcast) - if not readonly and array.flags.writeable: + if needs_writeable and not result.flags.writeable: result.flags.writeable = True return result	Error in broadcasting stride_tricks array We hit a new test failure when testing against numpy 1.10.1 : https://github.com/nipy/nibabel/pull/358 The test failure boils down to this: ``` import numpy as np shape = (2,) strides = [0] tricky_scalar = np.lib.stride_tricks.as_strided(np.array(0), shape, strides) other = np.zeros((1,)) first, second = np.broadcast_arrays(tricky_scalar, other) ``` On Linux (not OSX) I get the following error, for numpy 1.10.1: ``` Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array ``` There is no error for the same code on numpy 1.9.3. I think this is the same issue, arising in scipy: https://github.com/scipy/scipy/pull/5374#issuecomment-148775998	Hmm. I wrote the offending line of code, so I can take a look at this. @shoyer Any progress?	2015-10-24T19:01:58Z	[]	[]	Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array	10,403
numpy/numpy	numpy__numpy-6557	91093ff8c1ad9ccf8096fadc2e695d4039e529fc	diff --git a/numpy/lib/stride_tricks.py b/numpy/lib/stride_tricks.py --- a/numpy/lib/stride_tricks.py +++ b/numpy/lib/stride_tricks.py @@ -62,11 +62,14 @@ def _broadcast_to(array, shape, subok, readonly): if any(size < 0 for size in shape): raise ValueError('all elements of broadcast shape must be non-' 'negative') + needs_writeable = not readonly and array.flags.writeable + extras = ['reduce_ok'] if needs_writeable else [] + op_flag = 'readwrite' if needs_writeable else 'readonly' broadcast = np.nditer( - (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'], - op_flags=['readonly'], itershape=shape, order='C').itviews[0] + (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'] + extras, + op_flags=[op_flag], itershape=shape, order='C').itviews[0] result = _maybe_view_as_subclass(array, broadcast) - if not readonly and array.flags.writeable: + if needs_writeable and not result.flags.writeable: result.flags.writeable = True return result	Error in broadcasting stride_tricks array We hit a new test failure when testing against numpy 1.10.1 : https://github.com/nipy/nibabel/pull/358 The test failure boils down to this: ``` import numpy as np shape = (2,) strides = [0] tricky_scalar = np.lib.stride_tricks.as_strided(np.array(0), shape, strides) other = np.zeros((1,)) first, second = np.broadcast_arrays(tricky_scalar, other) ``` On Linux (not OSX) I get the following error, for numpy 1.10.1: ``` Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array ``` There is no error for the same code on numpy 1.9.3. I think this is the same issue, arising in scipy: https://github.com/scipy/scipy/pull/5374#issuecomment-148775998	Hmm. I wrote the offending line of code, so I can take a look at this. @shoyer Any progress?	2015-10-24T20:53:18Z	[]	[]	Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array	10,404
numpy/numpy	numpy__numpy-6905	e072d79f03610c33e336a9b700882d8905f9c958	diff --git a/numpy/lib/stride_tricks.py b/numpy/lib/stride_tricks.py --- a/numpy/lib/stride_tricks.py +++ b/numpy/lib/stride_tricks.py @@ -121,9 +121,6 @@ def _broadcast_shape(args): """ if not args: raise ValueError('must provide at least one argument') - if len(args) == 1: - # a single argument does not work with np.broadcast - return np.asarray(args[0]).shape # use the old-iterator because np.nditer does not handle size 0 arrays # consistently b = np.broadcast(args[:32])	Why does `numpy.broadcast` not accept a single input array? I'm wondering what's the reasoning behind forbidding a single array as input to `numpy.broadcast`. When writing code which is supposed to work in N dimensions with N >= 1, it would be nice to be able to use `broadcast` as-is. A standard use case is the calculation of the shape of an output array when using meshgrids as input for function evaluation: ``` python >>> coord_vecs = [[1, 2], [3, 4, 5]] >>> mesh = np.meshgrid(coord_vecs, indexing='ij', sparse=True) # for large stuff >>> out_shape = np.broadcast(mesh).shape >>> out_shape (2, 3) >>> coord_vecs = [[1, 2]] >>> mesh = np.meshgrid(coord_vecs, indexing='ij', sparse=True) # still works >>> out_shape = np.broadcast(mesh).shape Traceback (most recent call last): ... ValueError: Need at least two and fewer than (32) array objects. ``` As noted above, `meshgrid` works and returns a list with one element, as expected. Why does `broadcast` not return a broadcast object with shape `(2,)` in the example above? The other public methods of such an object would probably also make sense with a single input array. EDIT: Last sentence was wrong in the first version.	Apparently, there is [one single place](https://github.com/numpy/numpy/blob/004639d07fd161d1394f5dda1b6ed42c777f3c80/numpy/core/src/multiarray/iterators.c#L1606-L1614) where this situation is caught. Without knowing much about internals, the function around that check seems to be completely generic, and simply replacing 2 with 1 could be the only change necessary to allow single-array(-like) input. No idea about side effects of such a change, though. Well, maybe worth some playing around, then? I also find this surprising. Seems to work right away with the proposed (trivial) change: ``` python >>> import numpy as np >>> np.__path__ # My local GH copy ['/home/hkohr/Software/numpy/numpy'] >>> a = np.arange(3) >>> bc = np.broadcast(a) >>> bc.nd 1 >>> bc.numiter 1 >>> bc.shape (3,) >>> bc.size 3 ``` I get three errors when running `np.test()`, but they're probably unrelated. I just post them for the record: ``` ====================================================================== ERROR: Failure: ImportError (cannot import name ccompiler) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/loader.py", line 418, in loadTestsFromName addr.filename, addr.module) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/importer.py", line 47, in importFromPath return self.importFromDir(dir_path, fqname) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/importer.py", line 94, in importFromDir mod = load_module(part_fqname, fh, filename, desc) File "numpy/distutils/__init__.py", line 8, in <module> from . import ccompiler File "numpy/distutils/ccompiler.py", line 8, in <module> from distutils import ccompiler File "numpy/distutils/__init__.py", line 8, in <module> from . import ccompiler File "numpy/distutils/ccompiler.py", line 8, in <module> from distutils import ccompiler ImportError: cannot import name ccompiler ====================================================================== ERROR: test suite for <module 'test_array_from_pyobj' from '/home/hkohr/Software/numpy/numpy/f2py/tests/test_array_from_pyobj.py'> ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/suite.py", line 209, in run self.setUp() File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/suite.py", line 292, in setUp self.setupContext(ancestor) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/suite.py", line 315, in setupContext try_run(context, names) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/util.py", line 471, in try_run return func() File "/home/hkohr/Software/numpy/numpy/f2py/tests/test_array_from_pyobj.py", line 42, in setup 'test_array_from_pyobj_ext') File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 78, in wrapper memo[key] = func(a, kw) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 312, in build_module_distutils __import__(module_name) ImportError: /tmp/tmpqyEUs_/test_array_from_pyobj_ext.so: failed to map segment from shared object ====================================================================== ERROR: test_callback.TestF77Callback.test_string_callback ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/case.py", line 381, in setUp try_run(self.inst, ('setup', 'setUp')) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/util.py", line 471, in try_run return func() File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 361, in setUp module_name=self.module_name) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 78, in wrapper memo[key] = func(a, *kw) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 169, in build_code module_name=module_name) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 78, in wrapper memo[key] = func(a, **kw) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 149, in build_module __import__(module_name) ImportError: /tmp/tmpqyEUs_/_test_ext_module_5403.so: failed to map segment from shared object ---------------------------------------------------------------------- ``` +1 on exploring this change: looking at the code it certainly seems like an unnecessary restriction. Your test errors do not seem related, do you get the same errors if testing master unchanged? Would be good to put a PR together and see what Travis thinks.	2015-12-30T12:09:25Z	[]	[]	Traceback (most recent call last): ... ValueError: Need at least two and fewer than (32) array objects.	10,428
numpy/numpy	numpy__numpy-7133	e2805398f9a63b825f4a2aab22e9f169ff65aae9	diff --git a/numpy/lib/npyio.py b/numpy/lib/npyio.py --- a/numpy/lib/npyio.py +++ b/numpy/lib/npyio.py @@ -627,7 +627,11 @@ def _savez(file, args, kwds, compress, allow_pickle=True, pickle_kwargs=None): zipf = zipfile_factory(file, mode="w", compression=compression) # Stage arrays in a temporary file on disk, before writing to zip. - fd, tmpfile = tempfile.mkstemp(suffix='-numpy.npy') + + # Since target file might be big enough to exceed capacity of a global + # temporary directory, create temp file side-by-side with the target file. + file_dir, file_prefix = os.path.split(file) if _is_string_like(file) else (None, 'tmp') + fd, tmpfile = tempfile.mkstemp(prefix=file_prefix, dir=file_dir, suffix='-numpy.npy') os.close(fd) try: for key, val in namedict.items(): @@ -640,6 +644,8 @@ def _savez(file, args, kwds, compress, allow_pickle=True, pickle_kwargs=None): fid.close() fid = None zipf.write(tmpfile, arcname=fname) + except IOError as exc: + raise IOError("Failed to write to %s: %s" % (tmpfile, exc)) finally: if fid: fid.close()	Saving large array fails with savez_compressed() but works with save() I got this strange problem on a CentOS box, Python 2.6.6, Numpy 1.9.1: ``` [minhle@node069 ~]$ python Python 2.6.6 (r266:84292, Jan 22 2014, 09:42:36) [GCC 4.4.7 20120313 (Red Hat 4.4.7-4)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import numpy as np >>> np.version.version '1.9.1' >>> a = np.ones((775890380,)) >>> np.savez_compressed('/home/minhle/scratch/test.npz', a) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 560, in savez_compressed _savez(file, args, kwds, True) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 597, in _savez format.write_array(fid, np.asanyarray(val)) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/format.py", line 562, in write_array array.tofile(fp) IOError: 775890380 requested and 233691638 written >>> np.save('/home/minhle/scratch/test.npy', a) >>> b = np.load('/home/minhle/scratch/test.npy') >>> b[:10] array([ 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.]) >>> quit() [minhle@node069 ~]$ cat /etc/*-release CentOS release 6.5 (Final) Cluster Manager v5.2 slave LSB_VERSION=base-4.0-amd64:base-4.0-noarch:core-4.0-amd64:core-4.0-noarch:graphics-4.0-amd64:graphics-4.0-noarch:printing-4.0-amd64:printing-4.0-noarch CentOS release 6.5 (Final) CentOS release 6.5 (Final) ```	Got the same error... any solution in sight? I just got this error too but with `np.savez` as opposed to `np.savez_compressed`. For future reference, using `np.save` instead of `np.savez[_compressed]` seems to work as @omerlevy's example code indicates. So, is this an issue with `savez`? Anyone who knows numpy well have any ideas what's happening here? `savez` needs to store files temporarily on disk, and perhaps you run out of space on `/tmp`. Try setting `TMPDIR=$HOME/tmp` or so. @pv Yeah, changing `TMPDIR` did the trick for me. It may also be worth noting that I'm on an NFS filesystem. I'm not sure what sort of effects, if any, that might have. I'm guessing that I'm into subtleties in the way that /tmp is configured. However, I think numpy should make this issue explicit in the documentation and provide some semblance of an intelligent, approachable error message. Agreed. I ran onto this issue on HPC resources too and I agree that it would be good if the error / warning message recommended checking to make sure `TMPDIR` was set properly with adequate space. Ran into the same problem (kudos to ngoldbaum@IRC for pointing me here). I understand that for temp storage TEMPDIR is usually the logical location. BUT when dealing with large datasets, there is a reason why we might want to savez to a different location and /tmp would never be big enough. So why not to follow any other downloader's behavior and use targetfilename + '_tempSMTHRANDOM' suffix instead. This has higher possibility to succeed since would use the target location/partition which should have adequate amount of storage. Also error message should include location of that temp file to give users sensible feedback	2016-01-28T03:16:40Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 560, in savez_compressed _savez(file, args, kwds, True) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 597, in _savez format.write_array(fid, np.asanyarray(val)) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/format.py", line 562, in write_array array.tofile(fp) IOError: 775890380 requested and 233691638 written	10,431
numpy/numpy	numpy__numpy-7152	9cfdb218b16cba97800fc2ad0f37e1af780ab664	diff --git a/numpy/lib/arraypad.py b/numpy/lib/arraypad.py --- a/numpy/lib/arraypad.py +++ b/numpy/lib/arraypad.py @@ -1337,7 +1337,7 @@ def pad(array, pad_width, mode, **kwargs): 'reflect_type': 'even', } - if isinstance(mode, str): + if isinstance(mode, np.compat.basestring): # Make sure have allowed kwargs appropriate for mode for key in kwargs: if key not in allowedkwargs[mode]:	from __future__ import unicode_literals breaks numpy.pad The `pad` function accepts a `mode` parameter which can be a callable or str. [It uses the test `isinstance(mode, str)` to determine which is the case](https://github.com/numpy/numpy/blob/master/numpy/lib/arraypad.py#L1340). Using `from __future__ import unicode_literals` makes this fail, as the `isinstance` call returns `False`, eventually producing a `TypeError: 'unicode' object is not callable` when the `mode` argument is instead called as a function. For example, the script: ``` from __future__ import unicode_literals import numpy as np np.pad([10], 2, mode='constant') ``` Fails using `python2.7` with the error: ``` Traceback (most recent call last): File "test.py", line 4, in <module> np.pad([10], 2, mode='constant') File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/arraypad.py", line 1348, in pad kwargs) File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/shape_base.py", line 79, in apply_along_axis res = func1d(arr[tuple(i.tolist())],*args) TypeError: 'unicode' object is not callable ``` --- Would using the [`six.string_types`](https://pythonhosted.org/six/#six.string_types) comparison [_a la_ this stackexchange post](http://stackoverflow.com/a/11301392/230468) be a suitable fix?	Yeah, `np.pad` (and in general anything that takes a string `mode`-style argument) ought to accept unicode strings on py2. I don't know what the very simplest cleanest way to write this check would be, but we'd certainly accept a pull request fixing it. (Maybe a helper function like ``` def _py2_fixup_unicode(s): if this_is_py2 and isinstance(s, unicode): try: return s.encode("ascii") except UnicodeEncodeError: pass return s ``` and then functions with `mode`-style arguments do `mode = _fixup_py2_unicode(mode)` at the top?) It sounds like that might be the best, most-general solution. In which case, is `sys.version_info` a reliable way to check the python version? For `numpy.pad` in particular, since the only options are string and callable, what about using ``` if callable(mode): # treat as function.... else: # treat as str ``` That could also work, though since the specific thing we're worrying about is unicode-on-py2 I kinda prefer checking for unicode -- plus in your way, the code below has to be prepared to handle both unicode and str, while in mine it only has to work for str. But really there are lots of ways that would work -- someone just needs to write the patch :-) On Sun, Jan 24, 2016 at 2:50 PM, Luke notifications@github.com wrote: > It sounds like that might be the best, most-general solution. In which > case, is sys.version_info a reliable way to check the python version? > > For numpy.pad in particular, since the only options are string and > callable, what about using > > if callable(mode): > # treat as function.... > else: > # treat as str > > — > Reply to this email directly or view it on GitHub > https://github.com/numpy/numpy/issues/7112#issuecomment-174349793. ## Nathaniel J. Smith -- https://vorpus.org http://vorpus.org `np.compat.basestring` and/or `np.compat.sixu` might help? Anyway, I am not sure whether `np.pad` may be the least of our problems, do array fields and other C-side string arguments work with it? I don't remember fixing it, and I seem to remember it was a problem. Now that we have dropped Python 3.2 and 3.3, sixu can be replaced by `u"..."`. Unicode literals broke a lot of code, which is why they weren't part of the Python2, Python3 code unification, but it may be that we can do something about that now. I don't recall the exact problems that came up. The issue here isn't that we need Unicode literals but that np.pad should respond gracefully if someone passes in a Unicode literal. I agree that there are probably other places in the code that don't handle this well either (I think that's @seberg's point?), but I guess we should wait for a bug to be reported and then fix those too? On Jan 25, 2016 7:09 AM, "Charles Harris" notifications@github.com wrote: > Now that we have dropped Python 3.2 and 3.3, sixu can be replaced by > u"...". Unicode literals broke a lot of code, which is why they weren't > part of the Python2, Python3 code unification, but it may be that we can do > something about that now. I don't recall the exact problems that came up. > > — > Reply to this email directly or view it on GitHub > https://github.com/numpy/numpy/issues/7112#issuecomment-174537453. Yeah, as far as I understand Chuck we might now have a better chance to fix things then before. Array fields in dtypes definitely don't work in all regards at least. But that does not mean we should not just fix this. If we ever find something we cannot fix then so be it, but this should not be hard to fix.	2016-01-31T02:13:23Z	[]	[]	Traceback (most recent call last): File "test.py", line 4, in <module> np.pad([10], 2, mode='constant') File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/arraypad.py", line 1348, in pad kwargs) File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/shape_base.py", line 79, in apply_along_axis res = func1d(arr[tuple(i.tolist())],*args) TypeError: 'unicode' object is not callable	10,432
numpy/numpy	numpy__numpy-7587	1fc180b4c683e79649e5699303722995ca3e8ef9	diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -23,7 +23,7 @@ csingle, cdouble, inexact, complexfloating, newaxis, ravel, all, Inf, dot, add, multiply, sqrt, maximum, fastCopyAndTranspose, sum, isfinite, size, finfo, errstate, geterrobj, longdouble, rollaxis, amin, amax, product, abs, - broadcast, atleast_2d, intp, asanyarray, isscalar + broadcast, atleast_2d, intp, asanyarray, isscalar, object_ ) from numpy.lib import triu, asfarray from numpy.linalg import lapack_lite, _umath_linalg @@ -2112,7 +2112,7 @@ def norm(x, ord=None, axis=None, keepdims=False): """ x = asarray(x) - if not issubclass(x.dtype.type, inexact): + if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) # Immediately handle some default, simple, fast, and common cases.	Regression in linalg.norm() using dtype=object In NumPy 1.10.1, this works: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) array([ 0., 1.]) ``` In NumPy 1.11.0, however, it raises an exception: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence. ``` Probably this ``` python if not issubclass(x.dtype.type, inexact): x = x.astype(float) ``` should be changed to ``` python if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) ``` ? Does that make sense? If yes, I can make a PR.		2016-04-29T08:51:38Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence.	10,456
numpy/numpy	numpy__numpy-7608	730e2881219f4af79ab0ad5930a3bea5ab60b098	diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -23,7 +23,7 @@ csingle, cdouble, inexact, complexfloating, newaxis, ravel, all, Inf, dot, add, multiply, sqrt, maximum, fastCopyAndTranspose, sum, isfinite, size, finfo, errstate, geterrobj, longdouble, rollaxis, amin, amax, product, abs, - broadcast, atleast_2d, intp, asanyarray, isscalar + broadcast, atleast_2d, intp, asanyarray, isscalar, object_ ) from numpy.lib import triu, asfarray from numpy.linalg import lapack_lite, _umath_linalg @@ -2112,7 +2112,7 @@ def norm(x, ord=None, axis=None, keepdims=False): """ x = asarray(x) - if not issubclass(x.dtype.type, inexact): + if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) # Immediately handle some default, simple, fast, and common cases.	Regression in linalg.norm() using dtype=object In NumPy 1.10.1, this works: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) array([ 0., 1.]) ``` In NumPy 1.11.0, however, it raises an exception: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence. ``` Probably this ``` python if not issubclass(x.dtype.type, inexact): x = x.astype(float) ``` should be changed to ``` python if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) ``` ? Does that make sense? If yes, I can make a PR.		2016-05-07T01:56:25Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence.	10,458
numpy/numpy	numpy__numpy-8148	b0a6e5583f2952a040c45d0d50f5e93372dc175b	diff --git a/numpy/compat/py3k.py b/numpy/compat/py3k.py --- a/numpy/compat/py3k.py +++ b/numpy/compat/py3k.py @@ -118,7 +118,7 @@ def npy_load_module(name, fn, info=None): mod : module """ - import importlib + import importlib.machinery return importlib.machinery.SourceFileLoader(name, fn).load_module() else: def npy_load_module(name, fn, info=None):	Configuration.add_subpackage fails on Python 3.4 Using numpy 1.11.2 and Python 3.4, trying to run a `setupy.py` script which uses subpackage using numpy.distutil fails when `importlib.machinery` cannot be found in the `npy_load_module` function. For example using the following `setup.py` and `foo/setup.py` scripts: ``` Python # setup.py from numpy.distutils.misc_util import Configuration from numpy.distutils.core import setup def configuration(): config = Configuration(None, '', None) config.add_subpackage('foo') return config setup(configuration=configuration) ``` ``` Python # foo/setup.py def configuration(parent_package='', top_path=None): from numpy.distutils.misc_util import Configuration config = Configuration('foo', parent_package, top_path) return config ``` Attempting to build the package: ``` $ python setup.py build_ext -i Traceback (most recent call last): File "setup.py", line 10, in <module> setup(configuration=configuration) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/core.py", line 135, in setup config = configuration() File "setup.py", line 7, in configuration config.add_subpackage('foo') File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 1000, in add_subpackage caller_level = 2) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 969, in get_subpackage caller_level = caller_level + 1) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 882, in _get_configuration_from_setup_py ('.py', 'U', 1)) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/compat/py3k.py", line 112, in npy_load_module return importlib.machinery.SourceFileLoader(name, fn).load_module() AttributeError: 'module' object has no attribute 'machinery' ``` This works in Python 3.5 and 2.7 and with earlier versions of NumPy prior to the addition of `npy_load_module`. This bug can likely be fixed by adding `import importlib` to the `npy_load_module`. I will submit a PR shortly. A workaround is to `import importlib.machinery` somewhere in the root `setup.py` file.		2016-10-12T18:13:58Z	[]	[]	Traceback (most recent call last): File "setup.py", line 10, in <module> setup(configuration=configuration) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/core.py", line 135, in setup config = configuration() File "setup.py", line 7, in configuration config.add_subpackage('foo') File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 1000, in add_subpackage caller_level = 2) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 969, in get_subpackage caller_level = caller_level + 1) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 882, in _get_configuration_from_setup_py ('.py', 'U', 1)) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/compat/py3k.py", line 112, in npy_load_module return importlib.machinery.SourceFileLoader(name, fn).load_module() AttributeError: 'module' object has no attribute 'machinery'	10,492
numpy/numpy	numpy__numpy-8384	5f5ccecbfc116284ed8c8d53cd8b203ceef5f7c7	diff --git a/numpy/core/code_generators/genapi.py b/numpy/core/code_generators/genapi.py --- a/numpy/core/code_generators/genapi.py +++ b/numpy/core/code_generators/genapi.py @@ -469,7 +469,7 @@ def fullapi_hash(api_dicts): # To parse strings like 'hex = checksum' where hex is e.g. 0x1234567F and # checksum a 128 bits md5 checksum (hex format as well) -VERRE = re.compile('(^0x[\da-f]{8})\s=\s([\da-f]{32})') +VERRE = re.compile(r'(^0x[\da-f]{8})\s=\s([\da-f]{32})') def get_versions_hash(): d = [] diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -1644,7 +1644,7 @@ def gradient(f, varargs, kwargs): + \\left(h_{d}^{2} - h_{s}^{2}\\right)f\\left(x_{i}\\right) - h_{d}^{2}f\\left(x_{i}-h_{s}\\right)} { h_{s}h_{d}\\left(h_{d} + h_{s}\\right)} - + \mathcal{O}\\left(\\frac{h_{d}h_{s}^{2} + + \\mathcal{O}\\left(\\frac{h_{d}h_{s}^{2} + h_{s}h_{d}^{2}}{h_{d} + h_{s}}\\right) @@ -1656,7 +1656,7 @@ def gradient(f, varargs, *kwargs): \\hat f_{i}^{(1)}= \\frac{f\\left(x_{i+1}\\right) - f\\left(x_{i-1}\\right)}{2h} - + \mathcal{O}\\left(h^{2}\\right) + + \\mathcal{O}\\left(h^{2}\\right) With a similar procedure the forward/backward approximations used for boundaries can be derived. diff --git a/tools/find_deprecated_escaped_characters.py b/tools/find_deprecated_escaped_characters.py new file mode 100644 --- /dev/null +++ b/tools/find_deprecated_escaped_characters.py @@ -0,0 +1,69 @@ +#! /usr/bin/env python +""" +Look for escape sequences deprecated in Python 3.6. + +Python 3.6 deprecates a number of non-escape sequences starting with `\` that +were accepted before. For instance, '\(' was previously accepted but must now +be written as '\\(' or r'\('. + +""" +from __future__ import division, absolute_import, print_function + +import sys + +def main(root): + """Find deprecated escape sequences. + + Checks for deprecated escape sequences in ``.py files``. If `root` is a + file, that file is checked, if `root` is a directory all ``.py`` files + found in a recursive descent are checked. + + If a deprecated escape sequence is found, the file and line where found is + printed. Note that for multiline strings the line where the string ends is + printed and the error(s) are somewhere in the body of the string. + + Parameters + ---------- + root : str + File or directory to check. + Returns + ------- + None + + """ + count = 0 + + if sys.version_info[:2] >= (3, 6): + import ast + import tokenize + import warnings + from pathlib import Path + + base = Path(root) + paths = base.rglob(".py") if base.is_dir() else [base] + for path in paths: + # use tokenize to auto-detect encoding on systems where no + # default encoding is defined (e.g. LANG='C') + with tokenize.open(str(path)) as f: + with warnings.catch_warnings(record=True) as w: + warnings.simplefilter('always') + tree = ast.parse(f.read()) + if w: + print("file: ", str(path)) + for e in w: + print('line: ', e.lineno, ': ', e.message) + print() + count += len(w) + else: + raise RuntimeError("Python version must be >= 3.6") + + print("Errors Found", count) + + +if __name__ == "__main__": + from argparse import ArgumentParser + + parser = ArgumentParser(description="Find deprecated escaped characters") + parser.add_argument('root', help='directory or file to be checked') + args = parser.parse_args() + main(args.root)	`test_warning_calls` error on Python 3.6 As of today, the daily numpy wheel builds show the following error on master for all Python 3.6 builds: ``` ====================================================================== ERROR: test_warnings.test_warning_calls ---------------------------------------------------------------------- Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 81, in test_warning_calls tree = ast.parse(file.read()) File "/opt/cp36m/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 1675 SyntaxError: invalid escape sequence \m ``` https://travis-ci.org/MacPython/numpy-wheels/builds/204888912 This is testing commit 5f5ccecbf . I see that the usual travis-ci builds tested the same commit without error : https://travis-ci.org/numpy/numpy/builds/204805650 . Any thoughts?		2016-12-14T20:14:05Z	[]	[]	Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 81, in test_warning_calls tree = ast.parse(file.read()) File "/opt/cp36m/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 1675 SyntaxError: invalid escape sequence \m	10,504
numpy/numpy	numpy__numpy-8497	ee3ab365cb55cce6d0b9b6ed5cfbd8e3ede8cc66	diff --git a/numpy/matrixlib/defmatrix.py b/numpy/matrixlib/defmatrix.py --- a/numpy/matrixlib/defmatrix.py +++ b/numpy/matrixlib/defmatrix.py @@ -3,49 +3,15 @@ __all__ = ['matrix', 'bmat', 'mat', 'asmatrix'] import sys +import ast import numpy.core.numeric as N from numpy.core.numeric import concatenate, isscalar, binary_repr, identity, asanyarray from numpy.core.numerictypes import issubdtype -# make translation table -_numchars = '0123456789.-+jeEL' - -if sys.version_info[0] >= 3: - class _NumCharTable: - def __getitem__(self, i): - if chr(i) in _numchars: - return chr(i) - else: - return None - _table = _NumCharTable() - def _eval(astr): - str_ = astr.translate(_table) - if not str_: - raise TypeError("Invalid data string supplied: " + astr) - else: - return eval(str_) - -else: - _table = [None]*256 - for k in range(256): - _table[k] = chr(k) - _table = ''.join(_table) - - _todelete = [] - for k in _table: - if k not in _numchars: - _todelete.append(k) - _todelete = ''.join(_todelete) - del k - - def _eval(astr): - str_ = astr.translate(_table, _todelete) - if not str_: - raise TypeError("Invalid data string supplied: " + astr) - else: - return eval(str_) - def _convert_from_string(data): + for char in '[]': + data = data.replace(char, '') + rows = data.split(';') newdata = [] count = 0 @@ -54,7 +20,7 @@ def _convert_from_string(data): newrow = [] for col in trow: temp = col.split() - newrow.extend(map(_eval, temp)) + newrow.extend(map(ast.literal_eval, temp)) if count == 0: Ncols = len(newrow) elif len(newrow) != Ncols:	np.matrix('True False True') throws error When trying to create a numpy matrix using the string syntax `a = np.matrix('1 2; 3 4')` but with booleans instead of integers `np.matrix('True True False')` I see this error ``` >>> np.matrix('True True False') Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 267, in __new__ data = _convert_from_string(data) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 57, in _convert_from_string newrow.extend(map(_eval, temp)) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 26, in _eval return eval(str_) File "<string>", line 1, in <module> NameError: name 'e' is not defined ``` I'm not sure but is this desired behavior? Thanks!		2017-01-18T21:30:07Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 267, in __new__ data = _convert_from_string(data) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 57, in _convert_from_string newrow.extend(map(_eval, temp)) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 26, in _eval return eval(str_) File "<string>", line 1, in <module> NameError: name 'e' is not defined	10,510
numpy/numpy	numpy__numpy-8508	a621a2b700415a5c155546f9cb1f064c6099579e	diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -6130,8 +6130,11 @@ def __new__(self): def __array_finalize__(self, obj): return - def __array_wrap__(self, obj): - return self + def __array_prepare__(self, obj, context=None): + return self.view(MaskedArray).__array_prepare__(obj, context) + + def __array_wrap__(self, obj, context=None): + return self.view(MaskedArray).__array_wrap__(obj, context) def __str__(self): return str(masked_print_option._display)	__array_prepare__ produces bad shape for np.ma.masked A somewhat contrived example ```python >>> source = np.ma.masked # comes from user function, could be anything >>> source = np.asanyarray(source) # force to array, so we can try to copy the type >>> outarr = np.zeros((2, 2)) # allocate a raw ndarray for the result >>> bad = source.__array_prepare__(outarr) # prepare that raw array for operations >>> bad.shape (2, 2) >>> bad.data.shape () # uh oh >>> bad.mask.shape () # spaghettios ``` Which leads to failures like ```python >>> res.transpose((1, 0)) Traceback (most recent call last): File "<pyshell#78>", line 1, in <module> np.asanyarray(np.ma.masked).__array_prepare__(np.zeros((2, 2))).transpose((1, 0)) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 2509, in wrapped_method result = getattr(self._data, funcname)(args, *params) ValueError: axes don't match array ``` This come up when trying to invoke #8441 on masked arrays. Is this a bug, or an invalid use of `__array_prepare__`?	Related: ```python def assert_broadcasts(ufunc, a, b): expected = np.broadcast(a, b).shape actual = ufunc(a, b).shape assert expected == actual assert_broadcasts(np.add, [1, 2, 3], 1) assert_broadcasts(np.add, [1, 2, 3], np.ma.masked) # AssertionError ```	2017-01-20T18:16:20Z	[]	[]	Traceback (most recent call last): File "<pyshell#78>", line 1, in <module> np.asanyarray(np.ma.masked).__array_prepare__(np.zeros((2, 2))).transpose((1, 0)) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 2509, in wrapped_method result = getattr(self._data, funcname)(args, *params) ValueError: axes don't match array	10,511
numpy/numpy	numpy__numpy-8647	b097bd7ed4fa8e574414c2a0df8e50ac27bffa6d	diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -6991,44 +6991,42 @@ def where(condition, x=_NoValue, y=_NoValue): [6.0 -- 8.0]] """ - missing = (x is _NoValue, y is _NoValue).count(True) + # handle the single-argument case + missing = (x is _NoValue, y is _NoValue).count(True) if missing == 1: raise ValueError("Must provide both 'x' and 'y' or neither.") if missing == 2: - return filled(condition, 0).nonzero() - - # Both x and y are provided - - # Get the condition - fc = filled(condition, 0).astype(MaskType) - notfc = np.logical_not(fc) - - # Get the data - xv = getdata(x) - yv = getdata(y) - if x is masked: - ndtype = yv.dtype - elif y is masked: - ndtype = xv.dtype - else: - ndtype = np.find_common_type([xv.dtype, yv.dtype], []) - - # Construct an empty array and fill it - d = np.empty(fc.shape, dtype=ndtype).view(MaskedArray) - np.copyto(d._data, xv.astype(ndtype), where=fc) - np.copyto(d._data, yv.astype(ndtype), where=notfc) - - # Create an empty mask and fill it - mask = np.zeros(fc.shape, dtype=MaskType) - np.copyto(mask, getmask(x), where=fc) - np.copyto(mask, getmask(y), where=notfc) - mask \|= getmaskarray(condition) - - # Use d._mask instead of d.mask to avoid copies - d._mask = mask if mask.any() else nomask + return nonzero(condition) + + # we only care if the condition is true - false or masked pick y + cf = filled(condition, False) + xd = getdata(x) + yd = getdata(y) + + # we need the full arrays here for correct final dimensions + cm = getmaskarray(condition) + xm = getmaskarray(x) + ym = getmaskarray(y) + + # deal with the fact that masked.dtype == float64, but we don't actually + # want to treat it as that. + if x is masked and y is not masked: + xd = np.zeros((), dtype=yd.dtype) + xm = np.ones((), dtype=ym.dtype) + elif y is masked and x is not masked: + yd = np.zeros((), dtype=xd.dtype) + ym = np.ones((), dtype=xm.dtype) + + data = np.where(cf, xd, yd) + mask = np.where(cf, xm, ym) + mask = np.where(cm, np.ones((), dtype=mask.dtype), mask) + + # collapse the mask, for backwards compatibility + if mask.dtype == np.bool_ and not mask.any(): + mask = nomask - return d + return masked_array(data, mask=mask) def choose(indices, choices, out=None, mode='raise'):	BUG: np.ma.where does not broadcast correctly ```python >>> x = np.eye(3) >>> y = np.eye(3) >>> np.where([0, 1, 0], x, y) array([[ 1., 0., 0.], [ 0., 1., 0.], [ 0., 0., 1.]]) >>> np.ma.where([0, 1, 0], x, y) Traceback (most recent call last): File "<pyshell#24>", line 1, in <module> np.ma.where([0, 1, 0], x, y) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 6964, in where np.copyto(d._data, xv.astype(ndtype), where=fc) ValueError: could not broadcast input array from shape (3,3) into shape (3) ``` BUG: np.ma.where does not handle structured dtypes correctly ```python >>> dt = np.dtype([('a', int), ('b', int)]) >>> x = np.array([(1, 2), (3, 4)], dtype=dt) >>> np.where([0, 1], x, np.array((2, 3), dtype=dt)).dtype dtype([('a', '<i4'), ('b', '<i4')]) >>> np.ma.where([0, 1], x, np.array((2, 3), dtype=dt)).dtype dtype('O') ```		2017-02-20T15:20:04Z	[]	[]	Traceback (most recent call last): File "<pyshell#24>", line 1, in <module> np.ma.where([0, 1, 0], x, y) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 6964, in where np.copyto(d._data, xv.astype(ndtype), where=fc) ValueError: could not broadcast input array from shape (3,3) into shape (3)	10,522
numpy/numpy	numpy__numpy-8665	85cd7b7e1ef04a498dbb84fc7d7fb35881a73183	diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -6372,21 +6372,16 @@ def getdoc(self): def __call__(self, a, args, params): if self.reversed: args = list(args) - arr = args[0] - args[0] = a - a = arr - # Get the method from the array (if possible) + a, args[0] = args[0], a + + marr = asanyarray(a) method_name = self.__name__ - method = getattr(a, method_name, None) - if method is not None: - return method(args, *params) - # Still here ? Then a is not a MaskedArray - method = getattr(MaskedArray, method_name, None) - if method is not None: - return method(MaskedArray(a), args, *params) - # Still here ? OK, let's call the corresponding np function - method = getattr(np, method_name) - return method(a, args, *params) + method = getattr(type(marr), method_name, None) + if method is None: + # use the corresponding np function + method = getattr(np, method_name) + + return method(marr, args, **params) all = _frommethod('all') @@ -6535,9 +6530,7 @@ def compressed(x): Equivalent method. """ - if not isinstance(x, MaskedArray): - x = asanyarray(x) - return x.compressed() + return asanyarray(x).compressed() def concatenate(arrays, axis=0): @@ -7683,6 +7676,10 @@ def asanyarray(a, dtype=None): <class 'numpy.ma.core.MaskedArray'> """ + # workaround for #8666, to preserve identity. Ideally the bottom line + # would handle this for us. + if isinstance(a, MaskedArray) and (dtype is None or dtype == a.dtype): + return a return masked_array(a, dtype=dtype, copy=False, keep_mask=True, subok=True)	np.ma.count and np.ma.copy on list input The `np.ma.count` and `np.ma.copy` behave strangly if the input is a `list` or `tuple` (probably also other python builtins). ``` >>> np.ma.copy([1,2,3]) # unexpected behaviour [1, 2, 3] >>> np.copy([1,2,3]) # expected behaviour array([1, 2, 3]) >>> np.ma.count([1,2,3]) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Programming\Anaconda\envs\test\lib\site-packages\numpy\ma\core.py", line 6389, in __call__ return method(args, params) TypeError: count() takes exactly one argument (0 given) ``` The reason for this behaviour is that `list` has* a `count` and `copy` method and `np.ma._frommethod` tries to call it instead of `np.ma.MaskedArray.count` (and copy).		2017-02-22T10:22:33Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Programming\Anaconda\envs\test\lib\site-packages\numpy\ma\core.py", line 6389, in __call__ return method(args, *params) TypeError: count() takes exactly one argument (0 given)	10,524
numpy/numpy	numpy__numpy-8750	6a3edf3210b439a4d1a51acb4e01bac017697ee6	diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -1135,7 +1135,7 @@ def average(a, axis=None, weights=None, returned=False): wgt = wgt.swapaxes(-1, axis) scl = wgt.sum(axis=axis, dtype=result_dtype) - if (scl == 0.0).any(): + if np.any(scl == 0.0): raise ZeroDivisionError( "Weights sum to zero, can't be normalized")	np.average crashes for 1D decimal object array (now only without weights) I'm running macOS 10.12.3, Python 3.6.0, and NumPy 1.12.0. When using the decimal package with NumPy, the arrays are of type 'object'. While most operations are performed flawlessly, calling np.average() with one of these arrays throws an error if weights are not provided (`AttributeError: 'decimal.Decimal' object has no attribute 'dtype'`), and a different error if weights are provided (`AttributeError: 'bool' object has no attribute 'any'`). Contrarily, calling np.mean() executes without error. The source of the problem in both cases appears to be the assumption that the result of folding the input array results in a standard supported dtype, rather than being based on the operations available to the object type. Therefore, I suspect that this issue will extend to any types which support the requisite numeric operations but are not natively supported by NumPy. I resolved the weighted problem for my use case by changing the line `if (scl == 0.0).any()` (line 1138 in `lib/function_base.py`) to check if scl were an array first, and if not then removing the `any()` call, though I don't know if that solution is desirable or acceptable for the purposes of NumPy. Sample input: ``` import numpy as np import decimal as dc values = np.array([dc.Decimal(x) for x in range(10)]) weights = np.array([dc.Decimal(x) for x in range(10)]) weights /= weights.sum() print(np.mean(values)) print(np.average(values, weights=weights)) ``` Corresponding output: ``` 4.5 Traceback (most recent call last): File "bugreport.py", line 9, in <module> print(np.average(values, weights=weights)) File "/usr/local/lib/python3.6/site-packages/numpy/lib/function_base.py", line 1138, in average if (scl == 0.0).any(): AttributeError: 'bool' object has no attribute 'any' ```	Yes, this seems like a bug and certainly doesn't reflected the intended behavior as described in the docs. Decimal arrays should be acceptable as arguments for this function.	2017-03-06T20:07:37Z	[]	[]	Traceback (most recent call last): File "bugreport.py", line 9, in <module> print(np.average(values, weights=weights)) File "/usr/local/lib/python3.6/site-packages/numpy/lib/function_base.py", line 1138, in average if (scl == 0.0).any(): AttributeError: 'bool' object has no attribute 'any'	10,531
numpy/numpy	numpy__numpy-8762	485b099cd4b82d65dc38cb2b28c7119f003c76c4	diff --git a/numpy/lib/polynomial.py b/numpy/lib/polynomial.py --- a/numpy/lib/polynomial.py +++ b/numpy/lib/polynomial.py @@ -1036,17 +1036,47 @@ class poly1d(object): poly1d([ 1, -3, 2]) """ - coeffs = None - order = None - variable = None __hash__ = None - def __init__(self, c_or_r, r=0, variable=None): + @property + def coeffs(self): + """ The polynomial coefficients """ + return self._coeffs + + @property + def variable(self): + """ The name of the polynomial variable """ + return self._variable + + # calculated attributes + @property + def order(self): + """ The order or degree of the polynomial """ + return len(self._coeffs) - 1 + + @property + def roots(self): + """ The roots of the polynomial, where self(x) == 0 """ + return roots(self._coeffs) + + # alias attributes + r = roots + c = coef = coefficients = coeffs + o = order + + def __init__(self, c_or_r, r=False, variable=None): if isinstance(c_or_r, poly1d): - for key in c_or_r.__dict__.keys(): - self.__dict__[key] = c_or_r.__dict__[key] + self._variable = c_or_r._variable + self._coeffs = c_or_r._coeffs + + if set(c_or_r.__dict__) - set(self.__dict__): + msg = ("In the future extra properties will not be copied " + "across when constructing one poly1d from another") + warnings.warn(msg, FutureWarning, stacklevel=2) + self.__dict__.update(c_or_r.__dict__) + if variable is not None: - self.__dict__['variable'] = variable + self._variable = variable return if r: c_or_r = poly(c_or_r) @@ -1056,11 +1086,10 @@ def __init__(self, c_or_r, r=0, variable=None): c_or_r = trim_zeros(c_or_r, trim='f') if len(c_or_r) == 0: c_or_r = NX.array([0.]) - self.__dict__['coeffs'] = c_or_r - self.__dict__['order'] = len(c_or_r) - 1 + self._coeffs = c_or_r if variable is None: variable = 'x' - self.__dict__['variable'] = variable + self._variable = variable def __array__(self, t=None): if t: @@ -1199,29 +1228,17 @@ def __rdiv__(self, other): __rtruediv__ = __rdiv__ def __eq__(self, other): + if not isinstance(other, poly1d): + return NotImplemented if self.coeffs.shape != other.coeffs.shape: return False return (self.coeffs == other.coeffs).all() def __ne__(self, other): + if not isinstance(other, poly1d): + return NotImplemented return not self.__eq__(other) - def __setattr__(self, key, val): - raise ValueError("Attributes cannot be changed this way.") - - def __getattr__(self, key): - if key in ['r', 'roots']: - return roots(self.coeffs) - elif key in ['c', 'coef', 'coefficients']: - return self.coeffs - elif key in ['o']: - return self.order - else: - try: - return self.__dict__[key] - except KeyError: - raise AttributeError( - "'%s' has no attribute '%s'" % (self.__class__, key)) def __getitem__(self, val): ind = self.order - val @@ -1237,10 +1254,9 @@ def __setitem__(self, key, val): raise ValueError("Does not support negative powers.") if key > self.order: zr = NX.zeros(key-self.order, self.coeffs.dtype) - self.__dict__['coeffs'] = NX.concatenate((zr, self.coeffs)) - self.__dict__['order'] = key + self._coeffs = NX.concatenate((zr, self.coeffs)) ind = 0 - self.__dict__['coeffs'][ind] = val + self._coeffs[ind] = val return def __iter__(self):	numpy.poly1d.__eq__ method fails with AttributeError I called the `inspect.signature` function on a `numpy.poly1d` object: ```python import numpy as np import inspect poly = np.poly1d([1, 2, 3]) print(inspect.signature(poly)) ``` and got an unexpected exception: ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/python3.5/inspect.py", line 2987, in signature return Signature.from_callable(obj, follow_wrapped=follow_wrapped) File "/python3.5/inspect.py", line 2737, in from_callable follow_wrapper_chains=follow_wrapped) File "/python3.5/inspect.py", line 2228, in _signature_from_callable if _signature_is_builtin(obj): File "/python3.5/inspect.py", line 1785, in _signature_is_builtin obj in (type, object)) File "/python3.5/site-packages/numpy/lib/polynomial.py", line 1203, in __eq__ if self.coeffs.shape != other.coeffs.shape: AttributeError: type object 'type' has no attribute 'coeffs' ``` So `numpy.poly1d` objects can be compared only to objects with special attributes structure. The current`__eq__` implementation from [v.1.12.x](https://github.com/numpy/numpy/blob/maintenance/1.12.x/numpy/lib/polynomial.py#L1201-L1204): ```python def __eq__(self, other): if self.coeffs.shape != other.coeffs.shape: return False return (self.coeffs == other.coeffs).all() ``` numpy version 1.12.0	That should probably have a `if not isinstance(other, poly1d): return NotImplemented`. In fact, a whole bunch of the `__(.*)__` methods should have those lines I think `if not isinstance(other, poly1d): return False` is better. With `raise NotImplemented` we still will have the same problem. Is there any reason to not return False in such cases? > With raise NotImplemented we still will have the same problem. @bondarevts: Indeed, that is why I suggested `return NotImplemented` and not `raise NotImplementedError`. `NotImplemented` here just means "I have no idea, ask the other guy", and does not result in an error > Indeed, that is why I suggested return NotImplemented @eric-wieser You're right. I read it wrong. Thank you for the clarification.	2017-03-09T02:57:50Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/python3.5/inspect.py", line 2987, in signature return Signature.from_callable(obj, follow_wrapped=follow_wrapped) File "/python3.5/inspect.py", line 2737, in from_callable follow_wrapper_chains=follow_wrapped) File "/python3.5/inspect.py", line 2228, in _signature_from_callable if _signature_is_builtin(obj): File "/python3.5/inspect.py", line 1785, in _signature_is_builtin obj in (type, object)) File "/python3.5/site-packages/numpy/lib/polynomial.py", line 1203, in __eq__ if self.coeffs.shape != other.coeffs.shape: AttributeError: type object 'type' has no attribute 'coeffs'	10,532
numpy/numpy	numpy__numpy-8827	03f3789efe4da2c56d2841ed027ef6735ca2f11b	diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -19,12 +19,13 @@ import warnings from numpy.core import ( - array, asarray, zeros, empty, empty_like, transpose, intc, single, double, + array, asarray, zeros, empty, empty_like, intc, single, double, csingle, cdouble, inexact, complexfloating, newaxis, ravel, all, Inf, dot, add, multiply, sqrt, maximum, fastCopyAndTranspose, sum, isfinite, size, finfo, errstate, geterrobj, longdouble, moveaxis, amin, amax, product, abs, - broadcast, atleast_2d, intp, asanyarray, isscalar, object_, ones - ) + broadcast, atleast_2d, intp, asanyarray, isscalar, object_, ones, matmul, + swapaxes, divide) + from numpy.core.multiarray import normalize_axis_index from numpy.lib import triu, asfarray from numpy.linalg import lapack_lite, _umath_linalg @@ -223,6 +224,22 @@ def _assertNoEmpty2d(arrays): if _isEmpty2d(a): raise LinAlgError("Arrays cannot be empty") +def transpose(a): + """ + Transpose each matrix in a stack of matrices. + + Unlike np.transpose, this only swaps the last two axes, rather than all of + them + + Parameters + ---------- + a : (...,M,N) array_like + + Returns + ------- + aT : (...,N,M) ndarray + """ + return swapaxes(a, -1, -2) # Linear equations @@ -1279,7 +1296,7 @@ def eigh(a, UPLO='L'): # Singular value decomposition -def svd(a, full_matrices=1, compute_uv=1): +def svd(a, full_matrices=True, compute_uv=True): """ Singular Value Decomposition. @@ -1494,15 +1511,21 @@ def matrix_rank(M, tol=None): Rank of the array is the number of SVD singular values of the array that are greater than `tol`. + .. versionchanged:: 1.14 + Can now operate on stacks of matrices + Parameters ---------- M : {(M,), (..., M, N)} array_like input vector or stack of matrices - tol : {None, float}, optional - threshold below which SVD values are considered zero. If `tol` is - None, and ``S`` is an array with singular values for `M`, and - ``eps`` is the epsilon value for datatype of ``S``, then `tol` is - set to ``S.max() max(M.shape) * eps``. + tol : (...) array_like, float, optional + threshold below which SVD values are considered zero. If `tol` is + None, and ``S`` is an array with singular values for `M`, and + ``eps`` is the epsilon value for datatype of ``S``, then `tol` is + set to ``S.max() * max(M.shape) * eps``. + + .. versionchanged:: 1.14 + Broadcasted against the stack of matrices Notes ----- @@ -1569,6 +1592,8 @@ def matrix_rank(M, tol=None): S = svd(M, compute_uv=False) if tol is None: tol = S.max(axis=-1, keepdims=True) * max(M.shape[-2:]) * finfo(S.dtype).eps + else: + tol = asarray(tol)[...,newaxis] return (S > tol).sum(axis=-1) @@ -1582,26 +1607,29 @@ def pinv(a, rcond=1e-15 ): singular-value decomposition (SVD) and including all large singular values. + .. versionchanged:: 1.14 + Can now operate on stacks of matrices + Parameters ---------- - a : (M, N) array_like - Matrix to be pseudo-inverted. - rcond : float - Cutoff for small singular values. - Singular values smaller (in modulus) than - `rcond` * largest_singular_value (again, in modulus) - are set to zero. + a : (..., M, N) array_like + Matrix or stack of matrices to be pseudo-inverted. + rcond : (...) array_like of float + Cutoff for small singular values. + Singular values smaller (in modulus) than + `rcond` * largest_singular_value (again, in modulus) + are set to zero. Broadcasts against the stack of matrices Returns ------- - B : (N, M) ndarray - The pseudo-inverse of `a`. If `a` is a `matrix` instance, then so - is `B`. + B : (..., N, M) ndarray + The pseudo-inverse of `a`. If `a` is a `matrix` instance, then so + is `B`. Raises ------ LinAlgError - If the SVD computation does not converge. + If the SVD computation does not converge. Notes ----- @@ -1638,20 +1666,20 @@ def pinv(a, rcond=1e-15 ): """ a, wrap = _makearray(a) + rcond = asarray(rcond) if _isEmpty2d(a): res = empty(a.shape[:-2] + (a.shape[-1], a.shape[-2]), dtype=a.dtype) return wrap(res) a = a.conjugate() - u, s, vt = svd(a, 0) - m = u.shape[0] - n = vt.shape[1] - cutoff = rcondmaximum.reduce(s) - for i in range(min(n, m)): - if s[i] > cutoff: - s[i] = 1./s[i] - else: - s[i] = 0. - res = dot(transpose(vt), multiply(s[:, newaxis], transpose(u))) + u, s, vt = svd(a, full_matrices=False) + + # discard small singular values + cutoff = rcond[..., newaxis] amax(s, axis=-1, keepdims=True) + large = s > cutoff + s = divide(1, s, where=large, out=s) + s[~large] = 0 + + res = matmul(transpose(vt), multiply(s[..., newaxis], transpose(u))) return wrap(res) # Determinant @@ -1987,13 +2015,13 @@ def lstsq(a, b, rcond="warn"): resids = array([sum((ravel(bstar)[n:])2)], dtype=result_real_t) else: - x = array(transpose(bstar)[:n,:], dtype=result_t, copy=True) + x = array(bstar.T[:n,:], dtype=result_t, copy=True) if results['rank'] == n and m > n: if isComplexType(t): - resids = sum(abs(transpose(bstar)[n:,:])2, axis=0).astype( + resids = sum(abs(bstar.T[n:,:])2, axis=0).astype( result_real_t, copy=False) else: - resids = sum((transpose(bstar)[n:,:])2, axis=0).astype( + resids = sum((bstar.T[n:,:])**2, axis=0).astype( result_real_t, copy=False) st = s[:min(n, m)].astype(result_real_t, copy=True)	BUG: Linalg.pinv fails on stacks of matrices ```python >>> a = np.stack((np.eye(3),)*4, axis=0) >>> ai = np.linalg.inv(a) >>> assert (a == ai).all() >>> api = np.linalg.pinv(a) Traceback (most recent call last): File "<pyshell#9>", line 1, in <module> np.linalg.pinv(a) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\linalg\linalg.py", line 1668, in pinv if s[i] > cutoff: ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() ```		2017-03-25T00:01:50Z	[]	[]	Traceback (most recent call last): File "<pyshell#9>", line 1, in <module> np.linalg.pinv(a) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\linalg\linalg.py", line 1668, in pinv if s[i] > cutoff: ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()	10,536
numpy/numpy	numpy__numpy-9285	c6533b6c386dc0f4009e5f3c5c545dde4d1b48a4	diff --git a/numpy/distutils/ccompiler.py b/numpy/distutils/ccompiler.py --- a/numpy/distutils/ccompiler.py +++ b/numpy/distutils/ccompiler.py @@ -80,6 +80,7 @@ def _needs_build(obj, cc_args, extra_postargs, pp_opts): return False + def replace_method(klass, method_name, func): if sys.version_info[0] < 3: m = types.MethodType(func, None, klass) @@ -88,6 +89,25 @@ def replace_method(klass, method_name, func): m = lambda self, args, kw: func(self, args, **kw) setattr(klass, method_name, m) + +###################################################################### +## Method that subclasses may redefine. But don't call this method, +## it i private to CCompiler class and may return unexpected +## results if used elsewhere. So, you have been warned.. + +def CCompiler_find_executables(self): + """ + Does nothing here, but is called by the get_version method and can be + overridden by subclasses. In particular it is redefined in the `FCompiler` + class where more documentation can be found. + + """ + pass + + +replace_method(CCompiler, 'find_executables', CCompiler_find_executables) + + # Using customized CCompiler.spawn. def CCompiler_spawn(self, cmd, display=None): """ diff --git a/numpy/distutils/fcompiler/intel.py b/numpy/distutils/fcompiler/intel.py --- a/numpy/distutils/fcompiler/intel.py +++ b/numpy/distutils/fcompiler/intel.py @@ -57,7 +57,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-xhost -fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): @@ -123,7 +123,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): diff --git a/numpy/distutils/intelccompiler.py b/numpy/distutils/intelccompiler.py --- a/numpy/distutils/intelccompiler.py +++ b/numpy/distutils/intelccompiler.py @@ -19,7 +19,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe @@ -59,7 +59,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -m64 -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe	numpy 1.13.0 doesn't build with Intel compilers I'm running into the problem below when trying to build `numpy` 1.13.0 with Intel compilers; building earlier numpy versions (1.12.1, 1.11.1, 1.10.4, ...) with this approach works fine. ``` $ python setup.py build --compiler=intel --fcompiler=intelem building library "npymath" sources Found executable /path/to/icc Could not locate executable ecc Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables' ``` The `Could not locate executable ecc` seems to be a hint, since this message doesn't occur with older numpy versions where building like this does work: ``` ... building library "npymath" sources Found executable /path/to/icc customize IntelEM64TFCompiler Found executable /path/to/ifort customize IntelEM64TFCompiler using config C compiler: icc -fPIC -fp-model strict -O3 -fomit-frame-pointer -openmp ... ``` Why is `numpy` looking for `ecc` when it found `icc`, and not continuing to go and find `ifort` like it did before? A similar problem was reported by @Eric89GXL in #9101, but there the problem was that `icc` could not be found, that's clearly not the problem here. Seeing this with Python 2.7.13, Intel compilers 2017.1.132, Intel MKL 2017.1.132.	I get the same error. Commenting out the line `self.find_executables()` works as a work-around. This method seems to be undefined for CCompiler/IntelCCompiler. #8961 is the only change from 1.12.1 and I don't see how that would be related to this. #8961 has added exactly the line causing this (`v = self.get_version()`). So an alternate fix to #8961 suggested in #8941 was `-fopenmp`, which would not be version dependent. Might want to try that. Hi everybody, In regard to this change, in numpy 1.13.0 I've found using f2py and intel compilers I got the following compilation error: `mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' AttributeError: LooseVersion instance has no attribute 'split'` which can be solved changing the above line by: `mpopt = 'openmp' if v and int(v.version[0]) < 15 else 'qopenmp'` Hope this info can be useful, ACM Geez, version is a `LooseVersion` instance. Seems kinda useless given the lack of documentation of the class and the variations of version in the wild. That could be fixed by calling `str` on it I suppose. So there are two problems: * `LooseVersion` is not a string * missing `find_executables`, which only seems to be defined for fortran compilers @boegel Looks like "intel" is so generic that the Itanium platform is also being searched, weird. EDIT: I don't think the ecc message is relevant here.	2017-06-21T15:59:33Z	[]	[]	Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables'	10,556
numpy/numpy	numpy__numpy-9299	753e148e424e89da60390d7f1e15e65153da5aa0	diff --git a/numpy/distutils/ccompiler.py b/numpy/distutils/ccompiler.py --- a/numpy/distutils/ccompiler.py +++ b/numpy/distutils/ccompiler.py @@ -80,6 +80,7 @@ def _needs_build(obj, cc_args, extra_postargs, pp_opts): return False + def replace_method(klass, method_name, func): if sys.version_info[0] < 3: m = types.MethodType(func, None, klass) @@ -88,6 +89,25 @@ def replace_method(klass, method_name, func): m = lambda self, args, kw: func(self, args, **kw) setattr(klass, method_name, m) + +###################################################################### +## Method that subclasses may redefine. But don't call this method, +## it i private to CCompiler class and may return unexpected +## results if used elsewhere. So, you have been warned.. + +def CCompiler_find_executables(self): + """ + Does nothing here, but is called by the get_version method and can be + overridden by subclasses. In particular it is redefined in the `FCompiler` + class where more documentation can be found. + + """ + pass + + +replace_method(CCompiler, 'find_executables', CCompiler_find_executables) + + # Using customized CCompiler.spawn. def CCompiler_spawn(self, cmd, display=None): """ diff --git a/numpy/distutils/fcompiler/intel.py b/numpy/distutils/fcompiler/intel.py --- a/numpy/distutils/fcompiler/intel.py +++ b/numpy/distutils/fcompiler/intel.py @@ -57,7 +57,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-xhost -fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): @@ -123,7 +123,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): diff --git a/numpy/distutils/intelccompiler.py b/numpy/distutils/intelccompiler.py --- a/numpy/distutils/intelccompiler.py +++ b/numpy/distutils/intelccompiler.py @@ -19,7 +19,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe @@ -59,7 +59,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -m64 -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe	numpy 1.13.0 doesn't build with Intel compilers I'm running into the problem below when trying to build `numpy` 1.13.0 with Intel compilers; building earlier numpy versions (1.12.1, 1.11.1, 1.10.4, ...) with this approach works fine. ``` $ python setup.py build --compiler=intel --fcompiler=intelem building library "npymath" sources Found executable /path/to/icc Could not locate executable ecc Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables' ``` The `Could not locate executable ecc` seems to be a hint, since this message doesn't occur with older numpy versions where building like this does work: ``` ... building library "npymath" sources Found executable /path/to/icc customize IntelEM64TFCompiler Found executable /path/to/ifort customize IntelEM64TFCompiler using config C compiler: icc -fPIC -fp-model strict -O3 -fomit-frame-pointer -openmp ... ``` Why is `numpy` looking for `ecc` when it found `icc`, and not continuing to go and find `ifort` like it did before? A similar problem was reported by @Eric89GXL in #9101, but there the problem was that `icc` could not be found, that's clearly not the problem here. Seeing this with Python 2.7.13, Intel compilers 2017.1.132, Intel MKL 2017.1.132.	I get the same error. Commenting out the line `self.find_executables()` works as a work-around. This method seems to be undefined for CCompiler/IntelCCompiler. #8961 is the only change from 1.12.1 and I don't see how that would be related to this. #8961 has added exactly the line causing this (`v = self.get_version()`). So an alternate fix to #8961 suggested in #8941 was `-fopenmp`, which would not be version dependent. Might want to try that. Hi everybody, In regard to this change, in numpy 1.13.0 I've found using f2py and intel compilers I got the following compilation error: `mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' AttributeError: LooseVersion instance has no attribute 'split'` which can be solved changing the above line by: `mpopt = 'openmp' if v and int(v.version[0]) < 15 else 'qopenmp'` Hope this info can be useful, ACM Geez, version is a `LooseVersion` instance. Seems kinda useless given the lack of documentation of the class and the variations of version in the wild. That could be fixed by calling `str` on it I suppose. So there are two problems: * `LooseVersion` is not a string * missing `find_executables`, which only seems to be defined for fortran compilers @boegel Looks like "intel" is so generic that the Itanium platform is also being searched, weird. EDIT: I don't think the ecc message is relevant here.	2017-06-26T17:23:43Z	[]	[]	Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables'	10,557
numpy/numpy	numpy__numpy-9552	ae17d2c93dfac88cca9859d8b49490deb3991f41	diff --git a/numpy/doc/basics.py b/numpy/doc/basics.py --- a/numpy/doc/basics.py +++ b/numpy/doc/basics.py @@ -158,8 +158,8 @@ numpy provides with ``np.finfo(np.longdouble)``. NumPy does not provide a dtype with more precision than C -``long double``\s; in particular, the 128-bit IEEE quad precision -data type (FORTRAN's ``REAL16``\) is not available. +``long double``\\s; in particular, the 128-bit IEEE quad precision +data type (FORTRAN's ``REAL16``\\) is not available. For efficient memory alignment, ``np.longdouble`` is usually stored padded with zero bits, either to 96 or 128 bits. Which is more efficient	test_warning_calls "invalid escape sequence \s" on Python 3.6 daily wheel builds. New errors today (13 August), for Linux and OSX: ``` ====================================================================== ERROR: numpy.tests.test_warnings.test_warning_calls ---------------------------------------------------------------------- Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 79, in test_warning_calls tree = ast.parse(file.read()) File "/usr/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 184 SyntaxError: invalid escape sequence \s ``` https://travis-ci.org/MacPython/numpy-wheels/jobs/264023630 https://travis-ci.org/MacPython/numpy-wheels/jobs/264023631 https://travis-ci.org/MacPython/numpy-wheels/jobs/264023635	Was worried about that, the problem is ``` file: numpy/doc/basics.py line: 184 : invalid escape sequence \s ``` Came from documentation fixes. Probably need a raw string, although I would prefer making the files genuine rst. I should add that check to the tests...	2017-08-13T23:06:02Z	[]	[]	Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 79, in test_warning_calls tree = ast.parse(file.read()) File "/usr/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 184 SyntaxError: invalid escape sequence \s	10,567
open-mmlab/mmdetection	open-mmlab__mmdetection-10568	8822264e185df57250ac15bdbb86ac5a383e6520	diff --git a/demo/video_gpuaccel_demo.py b/demo/video_gpuaccel_demo.py --- a/demo/video_gpuaccel_demo.py +++ b/demo/video_gpuaccel_demo.py @@ -52,7 +52,9 @@ def prefetch_batch_input_shape(model: nn.Module, ori_wh: Tuple[int, test_pipeline = Compose(cfg.test_dataloader.dataset.pipeline) data = {'img': np.zeros((h, w, 3), dtype=np.uint8), 'img_id': 0} data = test_pipeline(data) - _, data_sample = model.data_preprocessor([data], False) + data['inputs'] = [data['inputs']] + data['data_samples'] = [data['data_samples']] + data_sample = model.data_preprocessor(data, False)['data_samples'] batch_input_shape = data_sample[0].batch_input_shape return batch_input_shape @@ -69,8 +71,8 @@ def pack_data(frame_resize: np.ndarray, batch_input_shape: Tuple[int, int], 'scale_factor': (batch_input_shape[0] / ori_shape[0], batch_input_shape[1] / ori_shape[1]) }) - frame_resize = torch.from_numpy(frame_resize).permute((2, 0, 1)) - data = {'inputs': frame_resize, 'data_sample': data_sample} + frame_resize = torch.from_numpy(frame_resize).permute((2, 0, 1)).cuda() + data = {'inputs': [frame_resize], 'data_samples': [data_sample]} return data @@ -112,7 +114,7 @@ def main(): for i, (frame_resize, frame_origin) in enumerate( zip(track_iter_progress(video_resize), video_origin)): data = pack_data(frame_resize, batch_input_shape, ori_shape) - result = model.test_step([data])[0] + result = model.test_step(data)[0] visualizer.add_datasample( name='video',	TypeError: list indices must be integers or slices, not str When I run the demo code video_gpuaccel_demo.py, it has the following error. How to solve it, thanks. Traceback (most recent call last): File "demo/video_gpuaccel_demo.py", line 147, in <module> main() File "demo/video_gpuaccel_demo.py", line 102, in main batch_input_shape = prefetch_batch_input_shape( File "demo/video_gpuaccel_demo.py", line 60, in prefetch_batch_input_shape _, data_sample = model.data_preprocessor([data], False) File "C:\Anaconda\Anaconda\envs\mmdetection\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(input, *kwargs) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 121, in forward batch_pad_shape = self._get_pad_shape(data) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 154, in _get_pad_shape _batch_inputs = data['inputs'] TypeError: list indices must be integers or slices, not str		2023-06-28T22:44:41Z	[]	[]	Traceback (most recent call last): File "demo/video_gpuaccel_demo.py", line 147, in <module> main() File "demo/video_gpuaccel_demo.py", line 102, in main batch_input_shape = prefetch_batch_input_shape( File "demo/video_gpuaccel_demo.py", line 60, in prefetch_batch_input_shape _, data_sample = model.data_preprocessor([data], False) File "C:\Anaconda\Anaconda\envs\mmdetection\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(input, *kwargs) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 121, in forward batch_pad_shape = self._get_pad_shape(data) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 154, in _get_pad_shape _batch_inputs = data['inputs'] TypeError: list indices must be integers or slices, not str	10,594
open-mmlab/mmdetection	open-mmlab__mmdetection-1099	b6712d4a9abe261b34b6a62f89ed3ed1fb88fae1	diff --git a/mmdet/core/bbox/__init__.py b/mmdet/core/bbox/__init__.py --- a/mmdet/core/bbox/__init__.py +++ b/mmdet/core/bbox/__init__.py @@ -1,4 +1,3 @@ -from .assign_sampling import assign_and_sample, build_assigner, build_sampler from .assigners import AssignResult, BaseAssigner, MaxIoUAssigner from .bbox_target import bbox_target from .geometry import bbox_overlaps @@ -9,6 +8,9 @@ bbox_mapping, bbox_mapping_back, delta2bbox, distance2bbox, roi2bbox) +from .assign_sampling import ( # isort:skip, avoid recursive imports + assign_and_sample, build_assigner, build_sampler) + __all__ = [ 'bbox_overlaps', 'BaseAssigner', 'MaxIoUAssigner', 'AssignResult', 'BaseSampler', 'PseudoSampler', 'RandomSampler',	ImportError: cannot import name 'build_sampler' from 'mmdet.core.bbox.assign_sampling' I have successful install the mmdetection by the command "pip install -v -e .". But I have the problem in the test. Would anyone help me ? (lab) gpuserver@ubuntu:~/ht/labs/mmdetection-master$ python Python 3.7.3 (default, Mar 27 2019, 22:11:17) [GCC 7.3.0] :: Anaconda, Inc. on linux Type "help", "copyright", "credits" or "license" for more information. >>> from mmdet.apis import init_detector Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/__init__.py", line 2, in <module> from .inference import inference_detector, init_detector, show_result File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/inference.py", line 9, in <module> from mmdet.core import get_classes File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/__init__.py", line 1, in <module> from .anchor import * # noqa: F401, F403 File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/__init__.py", line 2, in <module> from .anchor_target import anchor_inside_flags, anchor_target File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/anchor_target.py", line 3, in <module> from ..bbox import PseudoSampler, assign_and_sample, bbox2delta, build_assigner File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/__init__.py", line 1, in <module> from .assign_sampling import assign_and_sample, build_assigner, build_sampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py", line 3, in <module> from . import assigners, samplers File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/__init__.py", line 2, in <module> from .combined_sampler import CombinedSampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/combined_sampler.py", line 1, in <module> from ..assign_sampling import build_sampler ImportError: cannot import name 'build_sampler' from 'mmdet.core.bbox.assign_sampling' (/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py)	I encountered the same error today. same problem here.	2019-08-01T09:48:27Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/__init__.py", line 2, in <module> from .inference import inference_detector, init_detector, show_result File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/inference.py", line 9, in <module> from mmdet.core import get_classes File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/__init__.py", line 1, in <module> from .anchor import * # noqa: F401, F403 File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/__init__.py", line 2, in <module> from .anchor_target import anchor_inside_flags, anchor_target File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/anchor_target.py", line 3, in <module> from ..bbox import PseudoSampler, assign_and_sample, bbox2delta, build_assigner File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/__init__.py", line 1, in <module> from .assign_sampling import assign_and_sample, build_assigner, build_sampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py", line 3, in <module> from . import assigners, samplers File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/__init__.py", line 2, in <module> from .combined_sampler import CombinedSampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/combined_sampler.py", line 1, in <module> from ..assign_sampling import build_sampler ImportError: cannot import name 'build_sampler' from 'mmdet.core.bbox.assign_sampling' (/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py)	10,596
open-mmlab/mmdetection	open-mmlab__mmdetection-1404	c64beaf1494ea68562b274a723824d7f52fd31e8	diff --git a/mmdet/datasets/loader/sampler.py b/mmdet/datasets/loader/sampler.py --- a/mmdet/datasets/loader/sampler.py +++ b/mmdet/datasets/loader/sampler.py @@ -132,8 +132,12 @@ def __iter__(self): math.ceil( size * 1.0 / self.samples_per_gpu / self.num_replicas) ) * self.samples_per_gpu * self.num_replicas - len(indice) - indice += indice[:extra] - indices += indice + # pad indice + tmp = indice.copy() + for _ in range(extra // size): + indice.extend(tmp) + indice.extend(tmp[:extra % size]) + indices.extend(indice) assert len(indices) == self.total_size	assert len(indices) == self.total_size error during multiple GPU training I am trying to train my dataset on 8 GPU's. However, after calling `./dist_train.sh` this error assertion appeares: Traceback (most recent call last): File "./tools/train.py", line 113, in <module> main() File "./tools/train.py", line 109, in main logger=logger) File "/mmdetection/mmdet/apis/train.py", line 58, in train_detector _dist_train(model, dataset, cfg, validate=validate) File "/mmdetection/mmdet/apis/train.py", line 186, in _dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 358, in run epoch_runner(data_loaders[i], **kwargs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 260, in train for i, data_batch in enumerate(data_loader): File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 193, in __iter__ return _DataLoaderIter(self) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 493, in __init__ self._put_indices() File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 591, in _put_indices indices = next(self.sample_iter, None) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 172, in __iter__ for idx in self.sampler: File "/mmdetection/mmdet/datasets/loader/sampler.py", line 138, in __iter__ assert len(indices) == self.total_size ... in the config I tried various values for `imgs_per_gpu` and `workers_per_gpu`, currently it is: `imgs_per_gpu=2, workers_per_gpu=2,` no settings was working though. Single-GPU training works well. What is the meaning of this assert? Thanks!	Please follow the `Error report` issue template. > Please follow the `Error report` issue template. Here it is, thanks for any help! Checklist 1. I have searched related issues but cannot get the expected help. yes 2. The bug has not been fixed in the latest version. yup Describe the bug I am trying to train my custom dataset on 8 GPU's. However, after calling ./dist_train.sh the error showed below appeares. In the config I tried more values for `imgs_per_gpu` and `workers_per_gpu` (e.g. `imgs_per_gpu=2`, `workers_per_gpu=2`), no settings was working though. Single-GPU training works well. What is the meaning of the assert in the Traceback? What does not fit? Thanks! Reproduction 1. What command or script did you run? ``` ./tools/dist_train.sh MY_CONFIG 8 --validate ``` 2. Did you make any modifications on the code or config? Did you understand what you have modified? I modified number of classes, workers_per_gpu, imgs_per_gpu, dataset type and paths to the datasets. No changes in code. 3. What dataset did you use? My own dataset of 8 classes converted to COCO format. Environment - OS: Ubuntu 18.04 - GCC 5.4.0 - PyTorch version 1.1.0 - I built the docker I use on the official pytorch+cuda docker - GPU model: 8xV100 - CUDA version: 10.0, CUDNN version: 7.5 Error traceback ``` Traceback (most recent call last): File "./tools/train.py", line 113, in main() File "./tools/train.py", line 109, in main logger=logger) File "/mmdetection/mmdet/apis/train.py", line 58, in train_detector _dist_train(model, dataset, cfg, validate=validate) File "/mmdetection/mmdet/apis/train.py", line 186, in _dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 358, in run epoch_runner(data_loaders[i], kwargs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 260, in train for i, data_batch in enumerate(data_loader): File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 193, in iter return _DataLoaderIter(self) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 493, in init self._put_indices() File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 591, in _put_indices indices = next(self.sample_iter, None) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 172, in iter for idx in self.sampler: File "/mmdetection/mmdet/datasets/loader/sampler.py", line 138, in iter assert len(indices) == self.total_size ``` One more detail, when I print `len(indices)` and `self.total_size` right before the critical assert, it's `9308` and `9312`. The size of my training dataset is `9306`.. @hellock Any ideas? Recently I found out that when I set the config file to train with 2 GPUs (2 img/gpu), the training (called with `./dist_train`) initiates well. However, training with any more GPUs results in the mentioned assert error. Seems like a bug to me. You may count the images with aspect ratio >1 and <1. I suspect that there are only 2 images for one of the two groups. > You may count the images with aspect ratio >1 and <1. I suspect that there are only 2 images for one of the two groups. you are right, there are exactly 2 images with height > width The problem lies in [this line](https://github.com/open-mmlab/mmdetection/blob/master/mmdet/datasets/loader/sampler.py#L135) where `len(indices) < extra`. I meet the same issue, how to fix it? @ZhexuanZhou Before bug fixed, simplest way is make your img_per_gpu as power of 2. e.g. 2, 4, 8, 16 ... Also, make your gpu number as power of 2. This works for me. > @ZhexuanZhou > Before bug fixed, simplest way is make your img_per_gpu as power of 2. > e.g. 2, 4, 8, 16 ... > Also, make your gpu number as power of 2. > This works for me. This fix doesn't work for me, as is already mentioned in the bug description > Describe the bug** > I am trying to train my custom dataset on 8 GPU's. However, after calling ./dist_train.sh the error showed below appeares. In the config I tried more values for `imgs_per_gpu` and `workers_per_gpu` (e.g. `imgs_per_gpu=2`, `workers_per_gpu=2`), no settings was working though. The easiest work-around for me was to comment out the two asserts in the `sampler.py` :) @FilipLangr Yeah, that might cause some samples lost but not that harmful. Hello! I have situation, when 6 pics, where w>h. 33994 pics, when h>w. And 2 pics, where h==w. I have deleted h==w pics and have AssertionError: assert len(indices) == self.total_size anyway :( Than I deleted w>h pics and get another error: TypeError: 'NoneType' object is not subscriptable	2019-09-16T10:01:55Z	[]	[]	Traceback (most recent call last): File "./tools/train.py", line 113, in <module> main() File "./tools/train.py", line 109, in main logger=logger) File "/mmdetection/mmdet/apis/train.py", line 58, in train_detector _dist_train(model, dataset, cfg, validate=validate) File "/mmdetection/mmdet/apis/train.py", line 186, in _dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 358, in run epoch_runner(data_loaders[i], **kwargs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 260, in train for i, data_batch in enumerate(data_loader): File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 193, in __iter__ return _DataLoaderIter(self) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 493, in __init__ self._put_indices() File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 591, in _put_indices indices = next(self.sample_iter, None) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 172, in __iter__ for idx in self.sampler: File "/mmdetection/mmdet/datasets/loader/sampler.py", line 138, in __iter__ assert len(indices) == self.total_size ... in the config I tried various values for `imgs_per_gpu` and `workers_per_gpu`, currently it is:	10,598
open-mmlab/mmdetection	open-mmlab__mmdetection-2030	aade6801e7df66679b1fe9d162da0d03b4742dd4	diff --git a/mmdet/datasets/pipelines/transforms.py b/mmdet/datasets/pipelines/transforms.py --- a/mmdet/datasets/pipelines/transforms.py +++ b/mmdet/datasets/pipelines/transforms.py @@ -156,7 +156,7 @@ def _resize_masks(self, results): mmcv.imresize(mask, mask_size, interpolation='nearest') for mask in results[key] ] - results[key] = masks + results[key] = np.stack(masks) def _resize_seg(self, results): for key in results.get('seg_fields', []): @@ -245,10 +245,10 @@ def __call__(self, results): results['flip_direction']) # flip masks for key in results.get('mask_fields', []): - results[key] = [ + results[key] = np.stack([ mmcv.imflip(mask, direction=results['flip_direction']) for mask in results[key] - ] + ]) # flip segs for key in results.get('seg_fields', []): @@ -410,7 +410,7 @@ def __call__(self, results): gt_mask = results['gt_masks'][i][crop_y1:crop_y2, crop_x1:crop_x2] valid_gt_masks.append(gt_mask) - results['gt_masks'] = valid_gt_masks + results['gt_masks'] = np.stack(valid_gt_masks) return results @@ -586,7 +586,7 @@ def __call__(self, results): 0).astype(mask.dtype) expand_mask[top:top + h, left:left + w] = mask expand_gt_masks.append(expand_mask) - results['gt_masks'] = expand_gt_masks + results['gt_masks'] = np.stack(expand_gt_masks) # not tested if 'gt_semantic_seg' in results: @@ -678,10 +678,10 @@ def __call__(self, results): results['gt_masks'][i] for i in range(len(mask)) if mask[i] ] - results['gt_masks'] = [ + results['gt_masks'] = np.stack([ gt_mask[patch[1]:patch[3], patch[0]:patch[2]] for gt_mask in valid_masks - ] + ]) # not tested if 'gt_semantic_seg' in results:	Stacking of masks done in `Pad` Hey guys, I run into this error when commenting out the line ``` dict(type='Pad', size_divisor=32), ``` in `train_pipeline` on custom images. It causes this error: ``` Traceback (most recent call last): File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main "__main__", mod_spec) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/__main__.py", line 45, in <module> cli.main() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 361, in main run() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 203, in run_file runpy.run_path(options.target, run_name="__main__") File "/usr/lib/python3.6/runpy.py", line 263, in run_path pkg_name=pkg_name, script_name=fname) File "/usr/lib/python3.6/runpy.py", line 96, in _run_module_code mod_name, mod_spec, pkg_name, script_name) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/dev/mmdetection/tools/train.py", line 124, in <module> main() File "/home/pc/dev/mmdetection/tools/train.py", line 120, in main timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 133, in train_detector timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 319, in _non_dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 364, in run epoch_runner(data_loaders[i], kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 268, in train self.model, data_batch, train_mode=True, kwargs) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 100, in batch_processor losses = model(*data) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, *kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 150, in forward return self.module(inputs[0], *kwargs[0]) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, *kwargs) File "/home/pc/dev/mmdetection/mmdet/core/fp16/decorators.py", line 49, in new_func return old_func(args, kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/base.py", line 138, in forward return self.forward_train(img, img_meta, kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/two_stage.py", line 254, in forward_train sampling_results, gt_masks, self.train_cfg.rcnn) File "/home/pc/dev/mmdetection/mmdet/models/mask_heads/fcn_mask_head.py", line 111, in get_target gt_masks, rcnn_train_cfg) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 12, in mask_target mask_targets = torch.cat(list(mask_targets)) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 22, in mask_target_single _, maxh, maxw = gt_masks.shape AttributeError: 'list' object has no attribute 'shape' ``` I already debugged it and as far as I have found out it is because the masks are stacked in [`_pad_masks`](https://github.com/open-mmlab/mmdetection/blob/10c82efb0392fc1a5e1c696a53fe9ca7dfc3cdda/mmdet/datasets/pipelines/transforms.py#L304) If padding is not included in the pipeline, however, then `gt_masks` is a `list` instead of `ndarray` causing the above error. Shouldn't the stacking be done in [`_load_masks`](https://github.com/open-mmlab/mmdetection/blob/10c82efb0392fc1a5e1c696a53fe9ca7dfc3cdda/mmdet/datasets/pipelines/loading.py#L82) already to have a more flexible pipeline?	Thanks for the suggestions. It is a known issue and we are preparing a PR to fix it. Ok, thank you!	2020-01-31T04:36:49Z	[]	[]	Traceback (most recent call last): File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main "__main__", mod_spec) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/__main__.py", line 45, in <module> cli.main() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 361, in main run() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 203, in run_file runpy.run_path(options.target, run_name="__main__") File "/usr/lib/python3.6/runpy.py", line 263, in run_path pkg_name=pkg_name, script_name=fname) File "/usr/lib/python3.6/runpy.py", line 96, in _run_module_code mod_name, mod_spec, pkg_name, script_name) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/dev/mmdetection/tools/train.py", line 124, in <module> main() File "/home/pc/dev/mmdetection/tools/train.py", line 120, in main timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 133, in train_detector timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 319, in _non_dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 364, in run epoch_runner(data_loaders[i], kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 268, in train self.model, data_batch, train_mode=True, kwargs) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 100, in batch_processor losses = model(*data) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, *kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 150, in forward return self.module(inputs[0], *kwargs[0]) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, *kwargs) File "/home/pc/dev/mmdetection/mmdet/core/fp16/decorators.py", line 49, in new_func return old_func(args, kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/base.py", line 138, in forward return self.forward_train(img, img_meta, kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/two_stage.py", line 254, in forward_train sampling_results, gt_masks, self.train_cfg.rcnn) File "/home/pc/dev/mmdetection/mmdet/models/mask_heads/fcn_mask_head.py", line 111, in get_target gt_masks, rcnn_train_cfg) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 12, in mask_target mask_targets = torch.cat(list(mask_targets)) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 22, in mask_target_single _, maxh, maxw = gt_masks.shape AttributeError: 'list' object has no attribute 'shape'	10,603
open-mmlab/mmdetection	open-mmlab__mmdetection-2492	c1ef12df9c9d35f1402734435b23a4ae711f3084	diff --git a/mmdet/datasets/pipelines/transforms.py b/mmdet/datasets/pipelines/transforms.py --- a/mmdet/datasets/pipelines/transforms.py +++ b/mmdet/datasets/pipelines/transforms.py @@ -490,6 +490,9 @@ def __init__(self, def __call__(self, results): img = results['img'] + assert img.dtype == np.float32, \ + 'PhotoMetricDistortion needs the input image of dtype np.float32,'\ + ' please set "to_float32=True" in "LoadImageFromFile" pipeline' # random brightness if random.randint(2): delta = random.uniform(-self.brightness_delta,	_pickle.PicklingError: Can't pickle <class 'numpy.core._exceptions.UFuncTypeError'>: it's not the same object as numpy.core._exceptions.UFuncTypeError when i use dict(type='PhotoMetricDistortion', brightness_delta=32, contrast_range=(0.5, 1.5), saturation_range=(0.5, 1.5), hue_delta=18) in config file for training i got the following error: Traceback (most recent call last): File "/usr/lib/python3.6/multiprocessing/queues.py", line 234, in _feed obj = _ForkingPickler.dumps(obj) File "/usr/lib/python3.6/multiprocessing/reduction.py", line 51, in dumps cls(buf, protocol).dump(obj) _pickle.PicklingError: Can't pickle <class 'numpy.core._exceptions.UFuncTypeError'>: it's not the same object as numpy.core._exceptions.UFuncTypeError that's all info. i found the following code raise the error: delta = random.uniform(-self.brightness_delta,self.brightness_delta) img+=delta in PhotoMetricDistortion class in pipelines/transforms.py file other familiar code raise the error too. i don't know why? i found if delta is negative or float,it will raise the error.positive integer is ok.	when i use img = img + delta to replace img+=delta,it's ok... Do you use the config [ssd300_coco.py](https://github.com/open-mmlab/mmdetection/blob/master/configs/ssd300_coco.py), I can't reproduce your error with this config. Could you specify which config you use and what modification you apply to? Could you provide more detail information about your environment? when i use PhotoMetricDistortion in retinanet training, I met exactly same error > Do you use the config ssd300_coco.py, I can't reproduce your error with this config. Could you specify which config you use and what modification you apply to? Could you provide more detail information about your environment? I use the faster_rcnn_r50_fpn. torch-1.1.0, CUDA-10.1, numpy-1.18.2.Thanks. As mentioned above, when i replace '+=','=',and such operation in PhotoMetricDistortion, the error disappear. I found that the problem is that I didn't load image with float32=True, cv2.cvtColor doesn't support float64. That causes this wired error here:) > > Do you use the config ssd300_coco.py, I can't reproduce your error with this config. Could you specify which config you use and what modification you apply to? Could you provide more detail information about your environment? > > I use the faster_rcnn_r50_fpn. torch-1.1.0, CUDA-10.1, numpy-1.18.2.Thanks. As mentioned above, when i replace '+=','=',and such operation in PhotoMetricDistortion, the error disappear. @BChunlei Just as @edwardyangxin mentioned, when using PhotoMetricDistortion transformation, it is necessary to convert the image to np.float32 firstly just as what SSD does, you can refer [here](https://github.com/open-mmlab/mmdetection/blob/365c9302ee1eb5790b8e57f24ea4dfee8f2b88ac/configs/ssd300_coco.py#L51). But it is indeed not very friendly, would you like to create a PR to make it more friendly, i.e., report a reminder that user needs to set `to_float32` flag?	2020-04-20T14:48:29Z	[]	[]	Traceback (most recent call last): File "/usr/lib/python3.6/multiprocessing/queues.py", line 234, in _feed obj = _ForkingPickler.dumps(obj) File "/usr/lib/python3.6/multiprocessing/reduction.py", line 51, in dumps cls(buf, protocol).dump(obj) _pickle.PicklingError: Can't pickle <class 'numpy.core._exceptions.UFuncTypeError'>: it's not the same object as numpy.core._exceptions.UFuncTypeError	10,612
open-mmlab/mmdetection	open-mmlab__mmdetection-2824	24a43d5060adb1b523a018eeee17d5ca75b2b23a	diff --git a/mmdet/core/mask/__init__.py b/mmdet/core/mask/__init__.py --- a/mmdet/core/mask/__init__.py +++ b/mmdet/core/mask/__init__.py @@ -1,7 +1,8 @@ from .mask_target import mask_target from .structures import BitmapMasks, PolygonMasks -from .utils import split_combined_polys +from .utils import encode_mask_results, split_combined_polys __all__ = [ - 'split_combined_polys', 'mask_target', 'BitmapMasks', 'PolygonMasks' + 'split_combined_polys', 'mask_target', 'BitmapMasks', 'PolygonMasks', + 'encode_mask_results' ] diff --git a/mmdet/core/mask/utils.py b/mmdet/core/mask/utils.py --- a/mmdet/core/mask/utils.py +++ b/mmdet/core/mask/utils.py @@ -1,4 +1,6 @@ import mmcv +import numpy as np +import pycocotools.mask as mask_util def split_combined_polys(polys, poly_lens, polys_per_mask): @@ -28,3 +30,34 @@ def split_combined_polys(polys, poly_lens, polys_per_mask): mask_polys = mmcv.slice_list(split_polys, polys_per_mask_single) mask_polys_list.append(mask_polys) return mask_polys_list + + +# TODO: move this function to more proper place +def encode_mask_results(mask_results): + """Encode bitmap mask to RLE code. + + Args: + mask_results (list \| tuple[list]): bitmap mask results. + In mask scoring rcnn, mask_results is a tuple of (segm_results, + segm_cls_score). + + Returns: + list \| tuple: RLE encoded mask. + """ + if isinstance(mask_results, tuple): # mask scoring + cls_segms, cls_mask_scores = mask_results + else: + cls_segms = mask_results + num_classes = len(cls_segms) + encoded_mask_results = [[] for _ in range(num_classes)] + for i in range(len(cls_segms)): + for cls_segm in cls_segms[i]: + encoded_mask_results[i].append( + mask_util.encode( + np.array( + cls_segm[:, :, np.newaxis], order='F', + dtype='uint8'))[0]) # encoded with RLE + if isinstance(mask_results, tuple): + return encoded_mask_results, cls_mask_scores + else: + return encoded_mask_results diff --git a/mmdet/models/detectors/base.py b/mmdet/models/detectors/base.py --- a/mmdet/models/detectors/base.py +++ b/mmdet/models/detectors/base.py @@ -3,7 +3,6 @@ import mmcv import numpy as np -import pycocotools.mask as maskUtils import torch.nn as nn from mmcv.utils import print_log @@ -210,7 +209,7 @@ def show_result(self, for i in inds: i = int(i) color_mask = color_masks[labels[i]] - mask = maskUtils.decode(segms[i]).astype(np.bool) + mask = segms[i] img[mask] = img[mask] * 0.5 + color_mask * 0.5 # if out_file specified, do not show image in window if out_file is not None: diff --git a/mmdet/models/detectors/cascade_rcnn.py b/mmdet/models/detectors/cascade_rcnn.py --- a/mmdet/models/detectors/cascade_rcnn.py +++ b/mmdet/models/detectors/cascade_rcnn.py @@ -31,4 +31,4 @@ def show_result(self, data, result, kwargs): else: if isinstance(result, dict): result = result['ensemble'] - super(CascadeRCNN, self).show_result(data, result, kwargs) + return super(CascadeRCNN, self).show_result(data, result, **kwargs)	IndexError in pycocotools Thanks for your error report and we appreciate it a lot. Describe the bug When running the image_demo.py, I get an error about numpy index. Reproduction ``` python demo/image_demo.py demo/demo.jpg configs/mask_rcnn/mask_rcnn_r50_fpn_1x_coco.py checkpoints/mask_rcnn_r50_fpn_1x_coco_20200205-d4b0c5d6.pth --device cpu ``` Environment `ys.platform: linux Python: 3.7.4 (default, Aug 13 2019, 20:35:49) [GCC 7.3.0] CUDA available: True CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 10.0, V10.0.130 GPU 0: GeForce GTX 1080 Ti GCC: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609 PyTorch: 1.4.0 PyTorch compiling details: PyTorch built with: - GCC 7.3 - Intel(R) Math Kernel Library Version 2019.0.4 Product Build 20190411 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v0.21.1 (Git Hash 7d2fd500bc78936d1d648ca713b901012f470dbc) - OpenMP 201511 (a.k.a. OpenMP 4.5) - NNPACK is enabled - CUDA Runtime 10.1 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37 - CuDNN 7.6.3 - Magma 2.5.1 - Build settings: BLAS=MKL, BUILD_NAMEDTENSOR=OFF, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -O2 -fPIC -Wno-narrowing -Wall -Wextra -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Wno-stringop-overflow, DISABLE_NUMA=1, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_STATIC_DISPATCH=OFF, TorchVision: 0.5.0 OpenCV: 4.2.0 MMCV: 0.5.4 MMDetection: 2.0.0+unknown MMDetection Compiler: GCC 5.4 MMDetection CUDA Compiler: 10.0 Error traceback ``` Traceback (most recent call last): File "demo/image_demo.py", line 26, in <module> main() File "demo/image_demo.py", line 22, in main show_result_pyplot(model, args.img, result, score_thr=args.score_thr) File "/root/mmdetection-master/mmdet/apis/inference.py", line 146, in show_result_pyplot img = model.show_result(img, result, score_thr=score_thr, show=False) File "/root/mmdetection-master/mmdet/models/detectors/base.py", line 211, in show_result mask = maskUtils.decode(segms[i]).astype(np.bool) File "/root/anaconda3/lib/python3.7/site-packages/pycocotools-2.0-py3.7-linux-x86_64.egg/pycocotools/mask.py", line 91, in decode return _mask.decode([rleObjs])[:,:,0] File "pycocotools/_mask.pyx", line 146, in pycocotools._mask.decode File "pycocotools/_mask.pyx", line 128, in pycocotools._mask._frString IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices ```	Please try to install pycocotools through pip: `pip install "git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI"`. Sometimes the numpy version can also cause some problem. If the error still exist when you have installed the right pycocotools, you may try numpy=1.17 @Johnson-Wang I did install pycocotools from github, and the numpy version is 1.17, but the problem is not solved. @Johnson-Wang and I also tried lowering the numpy version, it just doesn't work Maybe try numpy>=1.18 ? @ZwwWayne Yes, I also try numpy 1.18, but it still doesn't work Did you install pycocotools from pip before you install from github using `pip install "git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI"`, it might because the environment already have one and the new installation does not work. In this case you need to uninstall all pycocotools in your environment and re-install a new one. @ZwwWayne Before install from github, I uninstalled pycocotools and make sure that there is no pycocotools libarary files in the system, but it just can't work. Then how about numpy? Did you do the same thing? Yes, I did everythingk, I even uninstall conda and clean all python environment, it doesn't work In object detection, try to rewrite the `def show_result(self, data, result, kwargs) `in your detector.py and add `self.CLASSES = tuple('A','B', ... , 'Z')` before `super(detector_name, self).show_result(data, result, kwargs)`. There may be similar labels in segm @aimhabo Actually I am doing something about instance segmentation, so I have to call the pycocotools , otherwise I will delete the segmentation part. But I don't understand how could similar labels cause an error in numpy? @mangdian I mean the similar ground-truth information like labels. In the dataset settings, it will load MS-COCO's information automatically while you use other dataset with `dataset_type = 'CocoDataset'`. In my similar situation, I force the detector's `self.CLASSES` in `show_result()` (In __init__() the forcement is failed because of any rewriting I haven't found yet). The problem is that the segmentation returned by the models is numpy boolean arrays instead of undecoded bytes data, so there is no need to decode the result. I finally solve this problem by modifying mmdet/apis/inference.py (delete the decode part). However, it is very strange that the code works well on another linux server of mine, no need to modify inference.py. Anyway, modifying inference.py can be a temporal solution to solve this problem. Hi @mangdian , Thanks for your bug report. Do you use the newest version when you meet the bug? It seems that we did not change inference.py after changing the test logic. Would you like to create a PR to fix that? @ZwwWayne Yes, I am using the newest version of mmdetection. What confuses me is that I tried the demo in two linux server. On one machine the demo code works, while on the other it doesn't work. I am not sure if it is a bug, or it is something related the numpy version or Cython version (if it is related to version, modifying the code may not be a good choice). But the reason that leads to the error is clear. The segmentation result returned by the model is numpy boolean array, insead of encoded bytes, so the cocotools failed to decode the results. It seems that some other people also meet this problem. Should I create a PR? Met the same problem here. I install the newest version mmdetection following the guide, it very annoying to met this error while testing`mask rcnn` with the demo code: ``` IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices ``` What's more, the model download urls for `aliyun` was not work @FantasyJXF Please try my solution: replace line 213 in mmdet/models/detectors/base.py: ` mask = maskUtils.decode(segms[i]).astype(np.bool) ` with `mask = segms[i]` > @FantasyJXF Please try my solution: > > replace line 213 in mmdet/models/detectors/base.py: > ` mask = maskUtils.decode(segms[i]).astype(np.bool)` > with > `mask = segms[i]` Very useful, thanks for your contribution. So there is no need to do the extra decode to get boolean array, the model already do that, maybe that's because the developer didn't check the segmentation model while release. First day using mmdetection, almost turn to dedectron2. Thank you again. @mangdian Thanks for the fix, it will be appreciated if you could create a PR for it. @hellock OK， I will create a PR for it. Glad to do it. Two things to fix: 1. Change https://github.com/open-mmlab/mmdetection/blob/master/mmdet/models/detectors/base.py#L213 to `mask = segms[i]`. 2. Change https://github.com/open-mmlab/mmdetection/blob/master/mmdet/models/detectors/cascade_rcnn.py#L34 to `return xxxx` #2697 @hellock Hi, anthor not related problem in [model zoo](https://github.com/open-mmlab/mmdetection/blob/master/docs/model_zoo.md) It seems the mirror sites for ALIYUN was not work, and the AWS mirror url postfix `open-mmlab` seems changed to `mmdetection`. Yes, the aliyun mirror site has not been updated to host v2.0 models. It will be updated later.	2020-05-27T02:14:48Z	[]	[]	Traceback (most recent call last): File "demo/image_demo.py", line 26, in <module> main() File "demo/image_demo.py", line 22, in main show_result_pyplot(model, args.img, result, score_thr=args.score_thr) File "/root/mmdetection-master/mmdet/apis/inference.py", line 146, in show_result_pyplot img = model.show_result(img, result, score_thr=score_thr, show=False) File "/root/mmdetection-master/mmdet/models/detectors/base.py", line 211, in show_result mask = maskUtils.decode(segms[i]).astype(np.bool) File "/root/anaconda3/lib/python3.7/site-packages/pycocotools-2.0-py3.7-linux-x86_64.egg/pycocotools/mask.py", line 91, in decode return _mask.decode([rleObjs])[:,:,0] File "pycocotools/_mask.pyx", line 146, in pycocotools._mask.decode File "pycocotools/_mask.pyx", line 128, in pycocotools._mask._frString IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices	10,617
open-mmlab/mmdetection	open-mmlab__mmdetection-3529	ae453fa92ffebcbd224b72f6d48e0b8699424450	diff --git a/tools/fuse_conv_bn.py b/tools/fuse_conv_bn.py deleted file mode 100644 --- a/tools/fuse_conv_bn.py +++ /dev/null @@ -1,67 +0,0 @@ -import argparse - -import torch -import torch.nn as nn -from mmcv.runner import save_checkpoint - -from mmdet.apis import init_detector - - -def fuse_conv_bn(conv, bn): - """During inference, the functionary of batch norm layers is turned off but - only the mean and var alone channels are used, which exposes the chance to - fuse it with the preceding conv layers to save computations and simplify - network structures.""" - conv_w = conv.weight - conv_b = conv.bias if conv.bias is not None else torch.zeros_like( - bn.running_mean) - - factor = bn.weight / torch.sqrt(bn.running_var + bn.eps) - conv.weight = nn.Parameter(conv_w * - factor.reshape([conv.out_channels, 1, 1, 1])) - conv.bias = nn.Parameter((conv_b - bn.running_mean) * factor + bn.bias) - return conv - - -def fuse_module(m): - last_conv = None - last_conv_name = None - - for name, child in m.named_children(): - if isinstance(child, (nn.BatchNorm2d, nn.SyncBatchNorm)): - if last_conv is None: # only fuse BN that is after Conv - continue - fused_conv = fuse_conv_bn(last_conv, child) - m._modules[last_conv_name] = fused_conv - # To reduce changes, set BN as Identity instead of deleting it. - m._modules[name] = nn.Identity() - last_conv = None - elif isinstance(child, nn.Conv2d): - last_conv = child - last_conv_name = name - else: - fuse_module(child) - return m - - -def parse_args(): - parser = argparse.ArgumentParser( - description='fuse Conv and BN layers in a model') - parser.add_argument('config', help='config file path') - parser.add_argument('checkpoint', help='checkpoint file path') - parser.add_argument('out', help='output path of the converted model') - args = parser.parse_args() - return args - - -def main(): - args = parse_args() - # build the model from a config file and a checkpoint file - model = init_detector(args.config, args.checkpoint) - # fuse conv and bn layers of the model - fused_model = fuse_module(model) - save_checkpoint(fused_model, args.out) - - -if __name__ == '__main__': - main()	ModuleNotFoundError: No module named 'tools' i would like to test the result of training, so i write the next: (base) zhangshen@zhangshen-X550JX:~/mmdetection$ python tools/test.py configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py checkpoints/faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth --out./result/result_100/pkl --eval bbox but i got: Traceback (most recent call last): File "tools/test.py", line 9, in <module> from tools.fuse_conv_bn import fuse_module ModuleNotFoundError: No module named 'tools' how can i solve this problem?	#2667. @yhcao6 Either fix the tool script or update the documentation. > @yhcao6 Either fix the tool script or update the documentation. how can i fix the tool script or which doucumentation should i update? I have created a pr to fix the error-report template, #3192 . Thanks for reporting the error.	2020-08-11T08:38:52Z	[]	[]	Traceback (most recent call last): File "tools/test.py", line 9, in <module> from tools.fuse_conv_bn import fuse_module ModuleNotFoundError: No module named 'tools'	10,626
open-mmlab/mmdetection	open-mmlab__mmdetection-3836	8e29b954a0276593fae3469eaf0d205da145c3da	diff --git a/mmdet/models/dense_heads/reppoints_head.py b/mmdet/models/dense_heads/reppoints_head.py --- a/mmdet/models/dense_heads/reppoints_head.py +++ b/mmdet/models/dense_heads/reppoints_head.py @@ -292,7 +292,7 @@ def forward_single(self, x): pts_out_refine = pts_out_refine + pts_out_init.detach() return cls_out, pts_out_init, pts_out_refine - def get_points(self, featmap_sizes, img_metas): + def get_points(self, featmap_sizes, img_metas, device): """Get points according to feature map sizes. Args: @@ -310,7 +310,7 @@ def get_points(self, featmap_sizes, img_metas): multi_level_points = [] for i in range(num_levels): points = self.point_generators[i].grid_points( - featmap_sizes[i], self.point_strides[i]) + featmap_sizes[i], self.point_strides[i], device) multi_level_points.append(points) points_list = [[point.clone() for point in multi_level_points] for _ in range(num_imgs)] @@ -326,7 +326,7 @@ def get_points(self, featmap_sizes, img_metas): valid_feat_h = min(int(np.ceil(h / point_stride)), feat_h) valid_feat_w = min(int(np.ceil(w / point_stride)), feat_w) flags = self.point_generators[i].valid_flags( - (feat_h, feat_w), (valid_feat_h, valid_feat_w)) + (feat_h, feat_w), (valid_feat_h, valid_feat_w), device) multi_level_flags.append(flags) valid_flag_list.append(multi_level_flags) @@ -534,6 +534,7 @@ def loss_single(self, cls_score, pts_pred_init, pts_pred_refine, labels, label_weights = label_weights.reshape(-1) cls_score = cls_score.permute(0, 2, 3, 1).reshape(-1, self.cls_out_channels) + cls_score = cls_score.contiguous() loss_cls = self.loss_cls( cls_score, labels, @@ -572,11 +573,12 @@ def loss(self, gt_bboxes_ignore=None): featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores] assert len(featmap_sizes) == len(self.point_generators) + device = cls_scores[0].device label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1 # target for initial stage center_list, valid_flag_list = self.get_points(featmap_sizes, - img_metas) + img_metas, device) pts_coordinate_preds_init = self.offset_to_pts(center_list, pts_preds_init) if self.train_cfg.init.assigner['type'] == 'PointAssigner': @@ -604,7 +606,7 @@ def loss(self, # target for refinement stage center_list, valid_flag_list = self.get_points(featmap_sizes, - img_metas) + img_metas, device) pts_coordinate_preds_refine = self.offset_to_pts( center_list, pts_preds_refine) bbox_list = [] @@ -666,6 +668,7 @@ def get_bboxes(self, rescale=False, nms=True): assert len(cls_scores) == len(pts_preds_refine) + device = cls_scores[0].device bbox_preds_refine = [ self.points2bbox(pts_pred_refine) for pts_pred_refine in pts_preds_refine @@ -673,7 +676,7 @@ def get_bboxes(self, num_levels = len(cls_scores) mlvl_points = [ self.point_generators[i].grid_points(cls_scores[i].size()[-2:], - self.point_strides[i]) + self.point_strides[i], device) for i in range(num_levels) ] result_list = []	RuntimeError: expected device cuda:1 but got device cuda:0 I have 2 titan xp gpus ,and when I run reppoints detection train.py it occurs: ``` python tools/train.py workproject/gureppoints/reppoints_moment_r101_fpn_gn-neck+head_2x_coco.py --gpu-ids=1 2020-09-24 17:04:35,811 - mmdet - INFO - Distributed training: False 2020-09-24 17:04:36,261 - mmdet - INFO - Config: dataset_type = 'CocoDataset' data_root = 'data/coco/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(2048, 1024), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(2048, 1024), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=4, train=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_train2017.json', img_prefix='data/coco/train2017/', pipeline=[ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(2048, 1024), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ]), val=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_val2017.json', img_prefix='data/coco/val2017/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(2048, 1024), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ]), test=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_test2017.json', img_prefix='data/coco/test2017/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(2048, 1024), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ])) evaluation = dict(interval=1, metric='bbox') optimizer = dict(type='SGD', lr=0.00125, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=None) lr_config = dict( policy='step', warmup='linear', warmup_iters=500, warmup_ratio=0.001, step=[16, 22]) total_epochs = 24 checkpoint_config = dict(interval=1) log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')]) dist_params = dict(backend='nccl') log_level = 'INFO' load_from = None resume_from = None workflow = [('train', 1)] model = dict( type='RepPointsDetector', pretrained='torchvision://resnet101', backbone=dict( type='ResNet', depth=101, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, style='pytorch'), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, start_level=1, add_extra_convs='on_input', num_outs=5, norm_cfg=dict(type='GN', num_groups=32, requires_grad=True)), bbox_head=dict( type='RepPointsHead', num_classes=6, in_channels=256, feat_channels=256, point_feat_channels=256, stacked_convs=3, num_points=9, gradient_mul=0.1, point_strides=[8, 16, 32, 64, 128], point_base_scale=4, loss_cls=dict( type='FocalLoss', use_sigmoid=True, gamma=2.0, alpha=0.25, loss_weight=1.0), loss_bbox_init=dict(type='SmoothL1Loss', beta=0.11, loss_weight=0.5), loss_bbox_refine=dict(type='SmoothL1Loss', beta=0.11, loss_weight=1.0), transform_method='moment', norm_cfg=dict(type='GN', num_groups=32, requires_grad=True))) train_cfg = dict( init=dict( assigner=dict(type='PointAssigner', scale=4, pos_num=1), allowed_border=-1, pos_weight=-1, debug=False), refine=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.4, min_pos_iou=0, ignore_iof_thr=-1), allowed_border=-1, pos_weight=-1, debug=False)) test_cfg = dict( nms_pre=1000, min_bbox_size=0, score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100) norm_cfg = dict(type='GN', num_groups=32, requires_grad=True) work_dir = 'workproject/gureppoints' gpu_ids = [1] 2020-09-24 17:04:37,362 - mmdet - INFO - load model from: torchvision://resnet101 2020-09-24 17:04:38,153 - mmdet - WARNING - The model and loaded state dict do not match exactly unexpected key in source state_dict: fc.weight, fc.bias loading annotations into memory... Done (t=0.29s) creating index... index created! loading annotations into memory... Done (t=0.06s) creating index... index created! 2020-09-24 17:04:44,095 - mmdet - INFO - Start running, host: ys@ys, work_dir: /media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/workproject/gureppoints 2020-09-24 17:04:44,096 - mmdet - INFO - workflow: [('train', 1)], max: 24 epochs Traceback (most recent call last): File "tools/train.py", line 178, in <module> main() File "tools/train.py", line 174, in main meta=meta) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/apis/train.py", line 143, in train_detector runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 122, in run epoch_runner(data_loaders[i], kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 32, in train kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(inputs[0], kwargs[0]) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 234, in train_step losses = self(data) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, *kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/core/fp16/decorators.py", line 51, in new_func return old_func(args, kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 168, in forward return self.forward_train(img, img_metas, kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/single_stage.py", line 94, in forward_train gt_labels, gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 581, in loss pts_preds_init) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 368, in offset_to_pts pts = xy_pts_shift self.point_strides[i_lvl] + pts_center RuntimeError: expected device cuda:1 but got device cuda:0 ``` so what's the problem?	Not typo. `print(args)` shows `gpu_ids=[1]`, and the log shows `gpu_ids = [1]`. > Not typo. > `print(args)` shows `gpu_ids=[1]`, and the log shows `gpu_ids = [1]`. you are right，I don't understand the usage of argparse well Does passing `device = 'cuda:1'` from `reppoints_head.py` work? `points = self.point_generators[i].grid_points(featmap_sizes[i], self.point_strides[i], device=device)` `flags = self.point_generators[i].valid_flags((feat_h, feat_w), (valid_feat_h, valid_feat_w), device=device)` If so, `pts = xy_pts_shift * self.point_strides[i_lvl] + pts_center.to(xy_pts_shift.device)` would be a tentative fix.	2020-09-25T07:15:13Z	[]	[]	Traceback (most recent call last): File "tools/train.py", line 178, in <module> main() File "tools/train.py", line 174, in main meta=meta) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/apis/train.py", line 143, in train_detector runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 122, in run epoch_runner(data_loaders[i], kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 32, in train kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(inputs[0], kwargs[0]) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 234, in train_step losses = self(data) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, *kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/core/fp16/decorators.py", line 51, in new_func return old_func(args, kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 168, in forward return self.forward_train(img, img_metas, kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/single_stage.py", line 94, in forward_train gt_labels, gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 581, in loss pts_preds_init) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 368, in offset_to_pts pts = xy_pts_shift self.point_strides[i_lvl] + pts_center RuntimeError: expected device cuda:1 but got device cuda:0	10,635
open-mmlab/mmdetection	open-mmlab__mmdetection-4555	3620bb74893ef688b3243652042660e1b5866d5e	diff --git a/mmdet/datasets/xml_style.py b/mmdet/datasets/xml_style.py --- a/mmdet/datasets/xml_style.py +++ b/mmdet/datasets/xml_style.py @@ -20,6 +20,8 @@ class XMLDataset(CustomDataset): """ def __init__(self, min_size=None, kwargs): + assert self.CLASSES or kwargs.get( + 'classes', None), 'CLASSES in `XMLDataset` can not be None.' super(XMLDataset, self).__init__(kwargs) self.cat2label = {cat: i for i, cat in enumerate(self.CLASSES)} self.min_size = min_size @@ -43,8 +45,6 @@ def load_annotations(self, ann_file): tree = ET.parse(xml_path) root = tree.getroot() size = root.find('size') - width = 0 - height = 0 if size is not None: width = int(size.find('width').text) height = int(size.find('height').text)	TypeError: argument of type 'NoneType' is not iterable When I use my own datasets to train faster-rcnn,I meet this question,the following are environment infomation and logs: sys.platform: linux Python: 3.7.9 \| packaged by conda-forge \| (default, Dec 9 2020, 21:08:20) [GCC 9.3.0] CUDA available: True GPU 0,1,2,3,4,5,6,7: GeForce GTX 1080 Ti CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 10.2, V10.2.89 GCC: gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 PyTorch: 1.5.0 PyTorch compiling details: PyTorch built with: - GCC 7.3 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v0.21.1 (Git Hash 7d2fd500bc78936d1d648ca713b901012f470dbc) - OpenMP 201511 (a.k.a. OpenMP 4.5) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_6 1;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37 - CuDNN 7.6.5 - Magma 2.5.2 - Build settings: BLAS=MKL, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -DNDEBUG -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DUSE_INTERNAL_THREADPOOL_IMPL -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_STATIC_DISPATCH=OFF, TorchVision: 0.6.0a0+82fd1c8 OpenCV: 4.4.0 MMCV: 1.2.4 MMCV Compiler: GCC 7.3 MMCV CUDA Compiler: 10.2 MMDetection: 2.8.0+ ------------------------------------------------------------ 2021-01-31 22:13:23,946 - mmdet - INFO - Distributed training: False 2021-01-31 22:13:26,538 - mmdet - INFO - Config: model = dict( type='FasterRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, style='pytorch'), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, feat_channels=256, anchor_generator=dict( type='AnchorGenerator', scales=[8], ratios=[0.5, 1.0, 2.0], strides=[4, 8, 16, 32, 64]), bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[1.0, 1.0, 1.0, 1.0]), loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0)), roi_head=dict( type='StandardRoIHead', bbox_roi_extractor=dict( type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0), out_channels=256, featmap_strides=[4, 8, 16, 32]), bbox_head=dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=232, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.1, 0.1, 0.2, 0.2]), reg_class_agnostic=False, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0))), train_cfg=dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, match_low_quality=True, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=-1, pos_weight=-1, debug=False), rpn_proposal=dict( nms_across_levels=False, nms_pre=2000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False)), test_cfg=dict( rpn=dict( nms_across_levels=False, nms_pre=1000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100))) dataset_type = 'VOCDataset' data_root = '/home/chengyuhong/mmdetection/data/tt100k_2021/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=2, train=dict( type='RepeatDataset', times=3, dataset=dict( type='VOCDataset', ann_file=[ '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/ImageSets/Main/trainval.txt' ], img_prefix=[ '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/' ], pipeline=[ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ])), val=dict( type='VOCDataset', ann_file= '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/ImageSets/Main/test.txt', img_prefix='/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ]), test=dict( type='VOCDataset', ann_file= '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/ImageSets/Main/test.txt', img_prefix='/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ])) evaluation = dict(interval=1, metric='mAP') checkpoint_config = dict(interval=1) log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')]) custom_hooks = [dict(type='NumClassCheckHook')] dist_params = dict(backend='nccl') log_level = 'INFO' load_from = None resume_from = None workflow = [('train', 1)] optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=None) lr_config = dict(policy='step', step=[3]) total_epochs = 4 work_dir = './work_dirs/faster_rcnn_r50_fpn_1x_voc0712' gpu_ids = range(0, 1) 2021-01-31 22:13:27,410 - mmdet - INFO - load model from: torchvision://resnet50 2021-01-31 22:13:27,935 - mmdet - WARNING - The model and loaded state dict do not match exactly unexpected key in source state_dict: fc.weight, fc.bias Traceback (most recent call last): File "./tools/train.py", line 187, in <module> main() File "./tools/train.py", line 163, in main datasets = [build_dataset(cfg.data.train)] File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 64, in build_dataset build_dataset(cfg['dataset'], default_args), cfg['times']) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 69, in build_dataset dataset = _concat_dataset(cfg, default_args) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 48, in _concat_dataset datasets.append(build_dataset(data_cfg, default_args)) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 71, in build_dataset dataset = build_from_cfg(cfg, DATASETS, default_args) File "/home/chengyuhong/anaconda3/envs/open-mmlab/lib/python3.7/site-packages/mmcv/utils/registry.py", line 171, in build_from_cfg return obj_cls(args) File "/home/chengyuhong/TT100K/mmdet/datasets/voc.py", line 32, in __init__ super(VOCDataset, self).__init__(kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 23, in __init__ super(XMLDataset, self).__init__(**kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/custom.py", line 96, in __init__ valid_inds = self._filter_imgs() File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 75, in _filter_imgs if name in self.CLASSES: TypeError: argument of type 'NoneType' is not iterable		2021-01-31T14:25:14Z	[]	[]	Traceback (most recent call last): File "./tools/train.py", line 187, in <module> main() File "./tools/train.py", line 163, in main datasets = [build_dataset(cfg.data.train)] File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 64, in build_dataset build_dataset(cfg['dataset'], default_args), cfg['times']) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 69, in build_dataset dataset = _concat_dataset(cfg, default_args) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 48, in _concat_dataset datasets.append(build_dataset(data_cfg, default_args)) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 71, in build_dataset dataset = build_from_cfg(cfg, DATASETS, default_args) File "/home/chengyuhong/anaconda3/envs/open-mmlab/lib/python3.7/site-packages/mmcv/utils/registry.py", line 171, in build_from_cfg return obj_cls(args) File "/home/chengyuhong/TT100K/mmdet/datasets/voc.py", line 32, in __init__ super(VOCDataset, self).__init__(kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 23, in __init__ super(XMLDataset, self).__init__(**kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/custom.py", line 96, in __init__ valid_inds = self._filter_imgs() File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 75, in _filter_imgs if name in self.CLASSES: TypeError: argument of type 'NoneType' is not iterable	10,650
open-mmlab/mmdetection	open-mmlab__mmdetection-5654	475c6bec197e6b495d48146636301590165d8f66	diff --git a/mmdet/models/detectors/detr.py b/mmdet/models/detectors/detr.py --- a/mmdet/models/detectors/detr.py +++ b/mmdet/models/detectors/detr.py @@ -1,3 +1,5 @@ +import warnings + import torch from ..builder import DETECTORS @@ -19,6 +21,27 @@ def __init__(self, super(DETR, self).__init__(backbone, None, bbox_head, train_cfg, test_cfg, pretrained, init_cfg) + # over-write `forward_dummy` because: + # the forward of bbox_head requires img_metas + def forward_dummy(self, img): + """Used for computing network flops. + + See `mmdetection/tools/analysis_tools/get_flops.py` + """ + warnings.warn('Warning! MultiheadAttention in DETR does not ' + 'support flops computation! Do not use the ' + 'results in your papers!') + + batch_size, _, height, width = img.shape + dummy_img_metas = [ + dict( + batch_input_shape=(height, width), + img_shape=(height, width, 3)) for _ in range(batch_size) + ] + x = self.extract_feat(img) + outs = self.bbox_head(x, dummy_img_metas) + return outs + # over-write `onnx_export` because: # (1) the forward of bbox_head requires img_metas # (2) the different behavior (e.g. construction of `masks`) between	Error get params DETR/ Deformable DETR Despite my attempts to modify, also just testing with the basic config detr file. Maybe this issue has already been raised? mmdet==2.13.0 mmcv=1.3.3 ```python python tools/analysis_tools/get_flops.py configs/detr/detr_r50_8x2_150e_coco.py ``` ```python /home/bluav/mmdetection/mmdet/models/backbones/resnet.py:400: UserWarning: DeprecationWarning: pretrained is a deprecated, please use "init_cfg" instead warnings.warn('DeprecationWarning: pretrained is a deprecated, ' Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(input, kwargs) File "/home/bluav/mmdetection/mmdet/models/detectors/single_stage.py", line 48, in forward_dummy outs = self.bbox_head(x) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(input, **kwargs) TypeError: forward() missing 1 required positional argument: 'img_metas' ```		2021-07-20T05:56:04Z	[]	[]	Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(input, kwargs) File "/home/bluav/mmdetection/mmdet/models/detectors/single_stage.py", line 48, in forward_dummy outs = self.bbox_head(x) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(input, **kwargs) TypeError: forward() missing 1 required positional argument: 'img_metas'	10,660
open-mmlab/mmdetection	open-mmlab__mmdetection-5884	6882fa0a220880e6e2e7e0536037f043b9031185	diff --git a/mmdet/models/roi_heads/bbox_heads/bbox_head.py b/mmdet/models/roi_heads/bbox_heads/bbox_head.py --- a/mmdet/models/roi_heads/bbox_heads/bbox_head.py +++ b/mmdet/models/roi_heads/bbox_heads/bbox_head.py @@ -455,9 +455,15 @@ def regress_by_class(self, rois, label, bbox_pred, img_meta): """Regress the bbox for the predicted class. Used in Cascade R-CNN. Args: - rois (Tensor): shape (n, 4) or (n, 5) - label (Tensor): shape (n, ) - bbox_pred (Tensor): shape (n, 4(#class)) or (n, 4) + rois (Tensor): Rois from `rpn_head` or last stage + `bbox_head`, has shape (num_proposals, 4) or + (num_proposals, 5). + label (Tensor): Only used when `self.reg_class_agnostic` + is False, has shape (num_proposals, ). + bbox_pred (Tensor): Regression prediction of + current stage `bbox_head`. When `self.reg_class_agnostic` + is False, it has shape (n, num_classes 4), otherwise + it has shape (n, 4). img_meta (dict): Image meta info. Returns: diff --git a/mmdet/models/roi_heads/cascade_roi_head.py b/mmdet/models/roi_heads/cascade_roi_head.py --- a/mmdet/models/roi_heads/cascade_roi_head.py +++ b/mmdet/models/roi_heads/cascade_roi_head.py @@ -280,7 +280,28 @@ def forward_train(self, return losses def simple_test(self, x, proposal_list, img_metas, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ assert self.with_bbox, 'Bbox head must be implemented.' num_imgs = len(proposal_list) img_shapes = tuple(meta['img_shape'] for meta in img_metas) @@ -340,7 +361,7 @@ def simple_test(self, x, proposal_list, img_metas, rescale=False): if rois[j].shape[0] > 0: bbox_label = cls_score[j][:, :-1].argmax(dim=1) refined_rois = self.bbox_head[i].regress_by_class( - rois[j], bbox_label[j], bbox_pred[j], img_metas[j]) + rois[j], bbox_label, bbox_pred[j], img_metas[j]) refine_rois_list.append(refined_rois) rois = torch.cat(refine_rois_list) diff --git a/mmdet/models/roi_heads/htc_roi_head.py b/mmdet/models/roi_heads/htc_roi_head.py --- a/mmdet/models/roi_heads/htc_roi_head.py +++ b/mmdet/models/roi_heads/htc_roi_head.py @@ -326,7 +326,28 @@ def forward_train(self, return losses def simple_test(self, x, proposal_list, img_metas, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ if self.with_semantic: _, semantic_feat = self.semantic_head(x) else: @@ -381,7 +402,7 @@ def simple_test(self, x, proposal_list, img_metas, rescale=False): if rois[j].shape[0] > 0: bbox_label = cls_score[j][:, :-1].argmax(dim=1) refine_rois = bbox_head.regress_by_class( - rois[j], bbox_label[j], bbox_pred[j], img_metas[j]) + rois[j], bbox_label, bbox_pred[j], img_metas[j]) refine_rois_list.append(refine_rois) rois = torch.cat(refine_rois_list) diff --git a/mmdet/models/roi_heads/scnet_roi_head.py b/mmdet/models/roi_heads/scnet_roi_head.py --- a/mmdet/models/roi_heads/scnet_roi_head.py +++ b/mmdet/models/roi_heads/scnet_roi_head.py @@ -213,26 +213,19 @@ def forward_train(self, """ Args: x (list[Tensor]): list of multi-level img features. - img_metas (list[dict]): list of image info dict where each dict has: 'img_shape', 'scale_factor', 'flip', and may also contain 'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'. For details on the values of these keys see `mmdet/datasets/pipelines/formatting.py:Collect`. - proposal_list (list[Tensors]): list of region proposals. - gt_bboxes (list[Tensor]): Ground truth bboxes for each image with shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format. - gt_labels (list[Tensor]): class indices corresponding to each box - gt_bboxes_ignore (None, list[Tensor]): specify which bounding boxes can be ignored when computing the loss. - gt_masks (None, Tensor) : true segmentation masks for each box used if the architecture supports a segmentation task. - gt_semantic_seg (None, list[Tensor]): semantic segmentation masks used if the architecture supports semantic segmentation task. @@ -317,7 +310,28 @@ def forward_train(self, return losses def simple_test(self, x, proposal_list, img_metas, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ if self.with_semantic: _, semantic_feat = self.semantic_head(x) else: @@ -379,7 +393,7 @@ def simple_test(self, x, proposal_list, img_metas, rescale=False): if rois[j].shape[0] > 0: bbox_label = cls_score[j][:, :-1].argmax(dim=1) refine_rois = bbox_head.regress_by_class( - rois[j], bbox_label[j], bbox_pred[j], img_metas[j]) + rois[j], bbox_label, bbox_pred[j], img_metas[j]) refine_rois_list.append(refine_rois) rois = torch.cat(refine_rois_list) diff --git a/mmdet/models/roi_heads/standard_roi_head.py b/mmdet/models/roi_heads/standard_roi_head.py --- a/mmdet/models/roi_heads/standard_roi_head.py +++ b/mmdet/models/roi_heads/standard_roi_head.py @@ -224,7 +224,28 @@ def simple_test(self, img_metas, proposals=None, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ assert self.with_bbox, 'Bbox head must be implemented.' det_bboxes, det_labels = self.simple_test_bboxes(	bbox_label dimension incorrect in cascade_roi_head.py simeple test Describe the bug I am trying the cascade rcnn network on a test dataset with a single image, after the training begins, the `regress_by_class` method called in `simple_test` of `cascade_roi_head.py` produce an error. Error traceback ``` 2021-08-07 12:48:22,062 - mmdet - INFO - Saving checkpoint at 1 epochs [ ] 0/1, elapsed: 0s Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 177, in main train_detector( File "c:\users\colli\mmdetection\mmdet\apis\train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 54, in train self.call_hook('after_train_epoch') File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\base_runner.py", line 307, in call_hook getattr(hook, fn_name)(self) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\hooks\evaluation.py", line 220, in after_train_epoch self._do_evaluate(runner) File "c:\users\colli\mmdetection\mmdet\core\evaluation\eval_hooks.py", line 17, in _do_evaluate results = single_gpu_test(runner.model, self.dataloader, show=False) File "c:\users\colli\mmdetection\mmdet\apis\test.py", line 27, in single_gpu_test result = model(return_loss=False, rescale=True, data) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(input, kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\parallel\data_parallel.py", line 42, in forward return super().forward(inputs, *kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\parallel\data_parallel.py", line 166, in forward return self.module(inputs[0], *kwargs[0]) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(input, *kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 98, in new_func return old_func(args, kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 173, in forward return self.forward_test(img, img_metas, kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 146, in forward_test return self.simple_test(imgs[0], img_metas[0], *kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\two_stage.py", line 181, in simple_test return self.roi_head.simple_test( File "c:\users\colli\mmdetection\mmdet\models\roi_heads\cascade_roi_head.py", line 344, in simple_test refined_rois = self.bbox_head[i].regress_by_class( File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 186, in new_func return old_func(args, kwargs) File "c:\users\colli\mmdetection\mmdet\models\roi_heads\bbox_heads\bbox_head.py", line 471, in regress_by_class inds = torch.stack((label, label + 1, label + 2, label + 3), 1) IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1) ``` Environment ``` sys.platform: win32 Python: 3.8.8 (default, Apr 13 2021, 15:08:03) [MSC v.1916 64 bit (AMD64)] CUDA available: True GPU 0: NVIDIA GeForce RTX 2060 SUPER CUDA_HOME: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.2 NVCC: Not Available GCC: n/a PyTorch: 1.9.0 PyTorch compiling details: PyTorch built with: - C++ Version: 199711 - MSVC 192829337 - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.1.2 (Git Hash 98be7e8afa711dc9b66c8ff3504129cb82013cdb) - OpenMP 2019 - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37 - CuDNN 7.6.5 - Magma 2.5.4 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=10.2, CUDNN_VERSION=7.6.5, CXX_COMPILER=C:/cb/pytorch_1000000000000/work/tmp_bin/sccache-cl.exe, CXX_FLAGS=/DWIN32 /D_WINDOWS /GR /EHsc /w /bigobj -DUSE_PTHREADPOOL -openmp:experimental -IC:/cb/pytorch_1000000000000/work/mkl/include -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.9.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=OFF, USE_NNPACK=OFF, USE_OPENMP=ON, TorchVision: 0.10.0 OpenCV: 4.5.2 MMCV: 1.3.9 MMCV Compiler: MSVC 192729111 MMCV CUDA Compiler: 11.2 MMDetection: 2.15.0+62a1cd3 ``` Bug fix** https://github.com/open-mmlab/mmdetection/blob/62a1cd3fd6091fc4efa83627631ab4b99a8d005c/mmdet/models/roi_heads/cascade_roi_head.py#L332-L345 https://github.com/open-mmlab/mmdetection/blob/46988b3ac9820bcd8728980f04be98272ee5ea39/mmdet/models/roi_heads/bbox_heads/bbox_head.py#L453-L465 `regress_by_class` expects a shape (n, ), but in the code above, in line 341 bbox_label is produced by slicing the cls_score, so it has been a 1 dimension tensor, but in line 343, it is sliced again, so it seems to be a zero dimension scalar now. I guess this produce the problem, after changing it to ``` refined_rois = self.bbox_head[i].regress_by_class(rois[j], bbox_label, bbox_pred[j], img_metas[j]) ``` The code runs correctly. The change in these lines is made in fd5d019ca983b032c35d15cd900b0fa6eec4f988 by @hhaAndroid, can you have a look at it? Am I understanding it correctly?	@hhaAndroid can you pls take a look at it? Can anyone have a look at this issue?	2021-08-14T03:18:04Z	[]	[]	Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 177, in main train_detector( File "c:\users\colli\mmdetection\mmdet\apis\train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 54, in train self.call_hook('after_train_epoch') File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\base_runner.py", line 307, in call_hook getattr(hook, fn_name)(self) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\hooks\evaluation.py", line 220, in after_train_epoch self._do_evaluate(runner) File "c:\users\colli\mmdetection\mmdet\core\evaluation\eval_hooks.py", line 17, in _do_evaluate results = single_gpu_test(runner.model, self.dataloader, show=False) File "c:\users\colli\mmdetection\mmdet\apis\test.py", line 27, in single_gpu_test result = model(return_loss=False, rescale=True, data) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(input, kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\parallel\data_parallel.py", line 42, in forward return super().forward(inputs, *kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\parallel\data_parallel.py", line 166, in forward return self.module(inputs[0], *kwargs[0]) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(input, *kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 98, in new_func return old_func(args, kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 173, in forward return self.forward_test(img, img_metas, kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 146, in forward_test return self.simple_test(imgs[0], img_metas[0], *kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\two_stage.py", line 181, in simple_test return self.roi_head.simple_test( File "c:\users\colli\mmdetection\mmdet\models\roi_heads\cascade_roi_head.py", line 344, in simple_test refined_rois = self.bbox_head[i].regress_by_class( File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 186, in new_func return old_func(args, **kwargs) File "c:\users\colli\mmdetection\mmdet\models\roi_heads\bbox_heads\bbox_head.py", line 471, in regress_by_class inds = torch.stack((label, label + 1, label + 2, label + 3), 1) IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1)	10,662
open-mmlab/mmdetection	open-mmlab__mmdetection-5930	0d2b39b195031b3b7ede73ebaa0a04150d1f332b	diff --git a/mmdet/datasets/pipelines/transforms.py b/mmdet/datasets/pipelines/transforms.py --- a/mmdet/datasets/pipelines/transforms.py +++ b/mmdet/datasets/pipelines/transforms.py @@ -2456,7 +2456,7 @@ def __call__(self, results): width = img.shape[1] + self.border[1] * 2 # Center - center_matrix = np.eye(3) + center_matrix = np.eye(3, dtype=np.float32) center_matrix[0, 2] = -img.shape[1] / 2 # x translation (pixels) center_matrix[1, 2] = -img.shape[0] / 2 # y translation (pixels) @@ -2561,21 +2561,24 @@ def __repr__(self): @staticmethod def _get_rotation_matrix(rotate_degrees): radian = math.radians(rotate_degrees) - rotation_matrix = np.array([[np.cos(radian), -np.sin(radian), 0.], - [np.sin(radian), - np.cos(radian), 0.], [0., 0., 1.]]) + rotation_matrix = np.array( + [[np.cos(radian), -np.sin(radian), 0.], + [np.sin(radian), np.cos(radian), 0.], [0., 0., 1.]], + dtype=np.float32) return rotation_matrix @staticmethod def _get_scaling_matrix(scale_ratio): - scaling_matrix = np.array([[scale_ratio, 0., 0.], - [0., scale_ratio, 0.], [0., 0., 1.]]) + scaling_matrix = np.array( + [[scale_ratio, 0., 0.], [0., scale_ratio, 0.], [0., 0., 1.]], + dtype=np.float32) return scaling_matrix @staticmethod def _get_share_matrix(scale_ratio): - scaling_matrix = np.array([[scale_ratio, 0., 0.], - [0., scale_ratio, 0.], [0., 0., 1.]]) + scaling_matrix = np.array( + [[scale_ratio, 0., 0.], [0., scale_ratio, 0.], [0., 0., 1.]], + dtype=np.float32) return scaling_matrix @staticmethod @@ -2583,10 +2586,12 @@ def _get_shear_matrix(x_shear_degrees, y_shear_degrees): x_radian = math.radians(x_shear_degrees) y_radian = math.radians(y_shear_degrees) shear_matrix = np.array([[1, np.tan(x_radian), 0.], - [np.tan(y_radian), 1, 0.], [0., 0., 1.]]) + [np.tan(y_radian), 1, 0.], [0., 0., 1.]], + dtype=np.float32) return shear_matrix @staticmethod def _get_translation_matrix(x, y): - translation_matrix = np.array([[1, 0., x], [0., 1, y], [0., 0., 1.]]) + translation_matrix = np.array([[1, 0., x], [0., 1, y], [0., 0., 1.]], + dtype=np.float32) return translation_matrix	RuntimeError: Expected object of scalar type Double but got scalar type Float for argument #2 'other' in call to _th_max when using mosaic mixup in cascade RCNN As mmdetection recently release 2.15.1, I want to use Mosaic and Mixup in Cascade RCNN, However, when I replace the relevant config from YOLOX to Cascade RCNN, I got the following error: ------------------------------------------------------------------------------------- 2021-08-12 06:53:24,512 - mmdet - INFO - workflow: [('train', 1)], max: 40 epochs Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 184, in main meta=meta) File "mmdetection-2.15.1/mmdet/apis/train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 50, in train self.run_iter(data_batch, train_mode=True, kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 30, in run_iter *kwargs) File "mmcv-1.3.9/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(inputs[0], kwargs[0]) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 237, in train_step losses = self(data) File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, kwargs) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 98, in new_func return old_func(args, kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 171, in forward return self.forward_train(img, img_metas, kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/two_stage.py", line 140, in forward_train proposal_cfg=proposal_cfg) File "mmdetection-2.15.1/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "mmdetection-2.15.1/mmdet/models/dense_heads/rpn_head.py", line 74, in loss gt_bboxes_ignore=gt_bboxes_ignore) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 186, in new_func return old_func(args, *kwargs) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 463, in loss label_channels=label_channels) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 345, in get_targets unmap_outputs=unmap_outputs) File "mmdetection-2.15.1/mmdet/core/utils/misc.py", line 29, in multi_apply return tuple(map(list, zip(map_results))) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 219, in _get_targets_single None if self.sampling else gt_labels) File "mmdetection-2.15.1/mmdet/core/bbox/assigners/max_iou_assigner.py", line 105, in assign overlaps = self.iou_calculator(gt_bboxes, bboxes) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 65, in __call__ return bbox_overlaps(bboxes1, bboxes2, mode, is_aligned) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 233, in bbox_overlaps bboxes2[..., None, :, :2]) # [B, rows, cols, 2] RuntimeError: Expected object of scalar type Double but got scalar type Float for argument #2 'other' in call to _th_max ------------------------------------------------------------------------------------- And below is my config: ------------------------------------------------------------------------------------- model = dict( type='CascadeRCNN', backbone=dict( type='ResNeXt', depth=101, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, style='pytorch', init_cfg=dict( type='Pretrained', checkpoint= 'pretrained_model/resnext101_64x4d-ee2c6f71.pth' ), groups=64, base_width=4), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, feat_channels=256, anchor_generator=dict( type='AnchorGenerator', scales=[8], ratios=[0.5, 1.0, 2.0], strides=[4, 8, 16, 32, 64]), bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[1.0, 1.0, 1.0, 1.0]), loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), loss_bbox=dict( type='SmoothL1Loss', beta=0.1111111111111111, loss_weight=1.0)), roi_head=dict( type='CascadeRoIHead', num_stages=3, stage_loss_weights=[1, 0.5, 0.25], bbox_roi_extractor=dict( type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0), out_channels=256, featmap_strides=[4, 8, 16, 32]), bbox_head=[ dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=12, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.1, 0.1, 0.2, 0.2]), reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)), dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=12, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.05, 0.05, 0.1, 0.1]), reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)), dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=12, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.033, 0.033, 0.067, 0.067]), reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)) ]), train_cfg=dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, match_low_quality=True, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=0, pos_weight=-1, debug=False), rpn_proposal=dict( nms_pre=2000, max_per_img=2000, nms=dict(type='nms', iou_threshold=0.7), min_bbox_size=0), rcnn=[ dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False), dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.6, neg_iou_thr=0.6, min_pos_iou=0.6, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False), dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.7, min_pos_iou=0.7, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False) ]), test_cfg=dict( rpn=dict( nms_pre=1000, max_per_img=1000, nms=dict(type='nms', iou_threshold=0.7), min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100))) dataset_type = 'CocoDataset' data_root = 'dataset/trainval/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) img_scale = (832, 832) train_pipeline = [ dict(type='Mosaic', img_scale=img_scale, pad_val=114.0), dict( type='RandomAffine', scaling_ratio_range=(0.1, 2), border=(-img_scale[0] // 2, -img_scale[1] // 2)), dict( type='MixUp', img_scale=img_scale, ratio_range=(0.8, 1.6), pad_val=114.0), dict( type='PhotoMetricDistortion', brightness_delta=32, contrast_range=(0.5, 1.5), saturation_range=(0.5, 1.5), hue_delta=18), dict(type='RandomFlip', flip_ratio=0.5), dict(type='Resize', keep_ratio=True), dict(type='Pad', pad_to_square=True, pad_val=114.0), dict(type='Normalize', img_norm_cfg), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=img_scale, flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict(type='Pad', size=img_scale, pad_val=114.0), dict(type='Normalize', img_norm_cfg), dict(type='ImageToTensor', keys=['img']), #try dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=8, workers_per_gpu=2, train=dict( type='MultiImageMixDataset', dataset=dict( type=dataset_type, ann_file= 'dataset/trainval/annotations/instances_train2017.json', img_prefix='dataset/trainval/', pipeline=[ dict(type='LoadImageFromFile', to_float32=True), dict(type='LoadAnnotations', with_bbox=True) ], filter_empty_gt=False, ), pipeline=train_pipeline, dynamic_scale=img_scale), val=dict( type=dataset_type, ann_file= 'dataset/trainval/annotations/instances_val2017.json', img_prefix='dataset/trainval/', pipeline=test_pipeline), test=dict( type=dataset_type, ann_file= 'dataset/trainval/annotations/instances_val2017.json', img_prefix='dataset/trainval/', ) ) evaluation = dict(interval=1, metric='bbox', save_best='bbox_mAP_50') optimizer = dict( type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005, nesterov=True, paramwise_cfg=dict(norm_decay_mult=0.0, bias_decay_mult=0.0)) optimizer_config = dict(grad_clip=None) lr_config = dict( policy='YOLOX', warmup='exp', by_epoch=False, warmup_by_epoch=True, warmup_ratio=1, warmup_iters=5, num_last_epochs=15, min_lr_ratio=0.05) runner = dict(type='EpochBasedRunner', max_epochs=40) checkpoint_config = dict(interval=1) log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')]) custom_hooks = [ dict(type='YOLOXModeSwitchHook', num_last_epochs=5, priority=48), dict( type='SyncRandomSizeHook', ratio_range=(14, 26), img_scale=(640, 640), interval=1, priority=48), dict(type='SyncNormHook', num_last_epochs=15, interval=1, priority=48), dict(type='ExpMomentumEMAHook', resume_from=None, priority=49) ] dist_params = dict(backend='nccl') log_level = 'INFO' load_from = None resume_from = None workflow = [('train', 1)] work_dir = './work_dirs/cascade_rcnn_x101_64x4d_fpn_20e_coco-3-832-mixupmosaic' gpu_ids = range(0, 1)	@jamiechoi1995 In fact, we did not test the above combination configuration, and there may be an incompatibility. > > > @jamiechoi1995 In fact, we did not test the above combination configuration, and there may be an incompatibility. I found that this error is due to the RandomAffine, MixUp and Mosaic augmentation returning Double type bbox, I solve it by forcing the bbox type of the above augmentation to float32. > > @jamiechoi1995 In fact, we did not test the above combination configuration, and there may be an incompatibility. > > I found that this error is due to the RandomAffine, MixUp and Mosaic augmentation returning Double type bbox, > I solve it by forcing the bbox type of the above augmentation to float32. Indeed possible. Can you create a PR to fix it? I find that it's because the dtype of warp_matrix in RandomAffine is float64, so after multiply this matrix to the box, the type of box becomes float64.	2021-08-23T10:42:16Z	[]	[]	Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 184, in main meta=meta) File "mmdetection-2.15.1/mmdet/apis/train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 50, in train self.run_iter(data_batch, train_mode=True, kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 30, in run_iter *kwargs) File "mmcv-1.3.9/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(inputs[0], kwargs[0]) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 237, in train_step losses = self(data) File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(input, kwargs) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 98, in new_func return old_func(args, kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 171, in forward return self.forward_train(img, img_metas, kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/two_stage.py", line 140, in forward_train proposal_cfg=proposal_cfg) File "mmdetection-2.15.1/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "mmdetection-2.15.1/mmdet/models/dense_heads/rpn_head.py", line 74, in loss gt_bboxes_ignore=gt_bboxes_ignore) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 186, in new_func return old_func(args, *kwargs) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 463, in loss label_channels=label_channels) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 345, in get_targets unmap_outputs=unmap_outputs) File "mmdetection-2.15.1/mmdet/core/utils/misc.py", line 29, in multi_apply return tuple(map(list, zip(map_results))) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 219, in _get_targets_single None if self.sampling else gt_labels) File "mmdetection-2.15.1/mmdet/core/bbox/assigners/max_iou_assigner.py", line 105, in assign overlaps = self.iou_calculator(gt_bboxes, bboxes) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 65, in __call__ return bbox_overlaps(bboxes1, bboxes2, mode, is_aligned) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 233, in bbox_overlaps bboxes2[..., None, :, :2]) # [B, rows, cols, 2] RuntimeError: Expected object of scalar type Double but got scalar type Float for argument #2 'other' in call to _th_max	10,663
open-mmlab/mmdetection	open-mmlab__mmdetection-6079	b1f97c1b6d2b3c735d8c5690952264c598c2f206	diff --git a/mmdet/models/dense_heads/yolact_head.py b/mmdet/models/dense_heads/yolact_head.py --- a/mmdet/models/dense_heads/yolact_head.py +++ b/mmdet/models/dense_heads/yolact_head.py @@ -663,6 +663,10 @@ def _init_layers(self): protonets = protonets[:-1] return nn.Sequential(*protonets) + def forward_dummy(self, x): + prototypes = self.protonet(x) + return prototypes + def forward(self, x, coeff_pred, bboxes, img_meta, sampling_results=None): """Forward feature from the upstream network to get prototypes and linearly combine the prototypes, using masks coefficients, into diff --git a/mmdet/models/detectors/yolact.py b/mmdet/models/detectors/yolact.py --- a/mmdet/models/detectors/yolact.py +++ b/mmdet/models/detectors/yolact.py @@ -30,7 +30,10 @@ def forward_dummy(self, img): See `mmdetection/tools/analysis_tools/get_flops.py` """ - raise NotImplementedError + feat = self.extract_feat(img) + bbox_outs = self.bbox_head(feat) + prototypes = self.mask_head.forward_dummy(feat[0]) + return (bbox_outs, prototypes) def forward_train(self, img,	Issue about tools\analysis_tools\get_flops.py When I used the tools/analysis_tools/get_flops.py to compute the Flops and Parameters of YOLACT , I got an error "NotImplementedError". Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 722, in _call_impl result = self.forward(input, *kwargs) File "/home/scsc01/lxx/mmlab/mmdet/models/detectors/yolact.py", line 32, in forward_dummy raise NotImplementedError NotImplementedError Does the get_flops.py not support the complexity calculation of YOLACT? thank you~~	In yolact.py, `forward_dummy` is not implemented https://github.com/open-mmlab/mmdetection/blob/6882fa0a220880e6e2e7e0536037f043b9031185/mmdet/models/detectors/yolact.py#L27-L32 But in get_flops.py https://github.com/open-mmlab/mmdetection/blob/6882fa0a220880e6e2e7e0536037f043b9031185/tools/analysis_tools/get_flops.py#L64-L69 I have checked `forward_dummy` of other models, and they are all implemented. I assume that the expected behavior is `forward_dummy` should not be declared in yolact.py, so the error information can be easier to understand? For now we have not supported it. We will support it in the future.	2021-09-10T13:27:44Z	[]	[]	Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 722, in _call_impl result = self.forward(input, *kwargs) File "/home/scsc01/lxx/mmlab/mmdet/models/detectors/yolact.py", line 32, in forward_dummy raise NotImplementedError NotImplementedError	10,665
open-mmlab/mmdetection	open-mmlab__mmdetection-7147	ffff556edc5a96ce72ce5b4d27d1fbcaa0d22122	diff --git a/tools/analysis_tools/confusion_matrix.py b/tools/analysis_tools/confusion_matrix.py --- a/tools/analysis_tools/confusion_matrix.py +++ b/tools/analysis_tools/confusion_matrix.py @@ -207,7 +207,10 @@ def plot_confusion_matrix(confusion_matrix, ax.text( j, i, - '{}%'.format(int(confusion_matrix[i, j])), + '{}%'.format( + int(confusion_matrix[ + i, + j]) if not np.isnan(confusion_matrix[i, j]) else -1), ha='center', va='center', color='w',	Confusion matrix error While trying to generate the confusion matrix with this command: ``` python tools/analysis_tools/confusion_matrix.py ./work_dirs/perception-types--D06-01-2022--T09-23-45/perception-types.py results.pkl ./temp --show ``` I ran into this error: ``` Traceback (most recent call last): File "tools/analysis_tools/confusion_matrix.py", line 261, in <module> main() File "tools/analysis_tools/confusion_matrix.py", line 257, in main show=args.show) File "tools/analysis_tools/confusion_matrix.py", line 210, in plot_confusion_matrix '{}%'.format(int(confusion_matrix[i, j])), ValueError: cannot convert float NaN to integer ``` Would appreciate any help or suggestions! Thanks	I ran into the same problem so I would appreciate if someone could help us with this. Thanks for reporting the bug. We will fix it asap.	2022-02-12T15:36:58Z	[]	[]	Traceback (most recent call last): File "tools/analysis_tools/confusion_matrix.py", line 261, in <module> main() File "tools/analysis_tools/confusion_matrix.py", line 257, in main show=args.show) File "tools/analysis_tools/confusion_matrix.py", line 210, in plot_confusion_matrix '{}%'.format(int(confusion_matrix[i, j])), ValueError: cannot convert float NaN to integer	10,676
open-mmlab/mmdetection	open-mmlab__mmdetection-7157	98949809b7179fab9391663ee5a4ab5978425f90	diff --git a/tools/deployment/onnx2tensorrt.py b/tools/deployment/onnx2tensorrt.py --- a/tools/deployment/onnx2tensorrt.py +++ b/tools/deployment/onnx2tensorrt.py @@ -201,7 +201,7 @@ def parse_args(): parsed directly from config file and are deprecated and will be \ removed in future releases.') if not args.input_img: - args.input_img = osp.join(osp.dirname(__file__), '../demo/demo.jpg') + args.input_img = osp.join(osp.dirname(__file__), '../../demo/demo.jpg') cfg = Config.fromfile(args.config)	FileNotFoundError: img file does not exist: tools/deployment/../demo/demo.jpg ``` Traceback (most recent call last): File "tools/deployment/onnx2tensorrt.py", line 255, in <module> verbose=args.verbose) File "tools/deployment/onnx2tensorrt.py", line 54, in onnx2tensorrt one_img, one_meta = preprocess_example_input(input_config) File "/workspace/mmdetection-2.18.1/mmdet/core/export/pytorch2onnx.py", line 139, in preprocess_example_input one_img = mmcv.imread(input_path) File "/workspace/mmcv-1.3.17-trt/mmcv/image/io.py", line 177, in imread f'img file does not exist: {img_or_path}') File "/workspace/mmcv-1.3.17-trt/mmcv/utils/path.py", line 23, in check_file_exist raise FileNotFoundError(msg_tmpl.format(filename)) FileNotFoundError: img file does not exist: tools/deployment/../demo/demo.jpg ```		2022-02-14T12:37:02Z	[]	[]	Traceback (most recent call last): File "tools/deployment/onnx2tensorrt.py", line 255, in <module> verbose=args.verbose) File "tools/deployment/onnx2tensorrt.py", line 54, in onnx2tensorrt one_img, one_meta = preprocess_example_input(input_config) File "/workspace/mmdetection-2.18.1/mmdet/core/export/pytorch2onnx.py", line 139, in preprocess_example_input one_img = mmcv.imread(input_path) File "/workspace/mmcv-1.3.17-trt/mmcv/image/io.py", line 177, in imread f'img file does not exist: {img_or_path}') File "/workspace/mmcv-1.3.17-trt/mmcv/utils/path.py", line 23, in check_file_exist raise FileNotFoundError(msg_tmpl.format(filename)) FileNotFoundError: img file does not exist: tools/deployment/../demo/demo.jpg	10,677
open-mmlab/mmdetection	open-mmlab__mmdetection-7407	c546b5044098b71d59a139036a87c5c97bcab4e2	diff --git a/tools/analysis_tools/analyze_logs.py b/tools/analysis_tools/analyze_logs.py old mode 100644 new mode 100755 --- a/tools/analysis_tools/analyze_logs.py +++ b/tools/analysis_tools/analyze_logs.py @@ -17,6 +17,10 @@ def cal_train_time(log_dicts, args): all_times.append(log_dict[epoch]['time']) else: all_times.append(log_dict[epoch]['time'][1:]) + if not all_times: + raise KeyError( + 'Please reduce the log interval in the config so that' + 'interval is less than iterations of one epoch.') all_times = np.array(all_times) epoch_ave_time = all_times.mean(-1) slowest_epoch = epoch_ave_time.argmax() @@ -50,12 +54,21 @@ def plot_curve(log_dicts, args): epochs = list(log_dict.keys()) for j, metric in enumerate(metrics): print(f'plot curve of {args.json_logs[i]}, metric is {metric}') - if metric not in log_dict[epochs[0]]: + if metric not in log_dict[epochs[int(args.start_epoch) - 1]]: + if 'mAP' in metric: + raise KeyError( + f'{args.json_logs[i]} does not contain metric ' + f'{metric}. Please check if "--no-validate" is ' + 'specified when you trained the model.') raise KeyError( - f'{args.json_logs[i]} does not contain metric {metric}') + f'{args.json_logs[i]} does not contain metric {metric}. ' + 'Please reduce the log interval in the config so that ' + 'interval is less than iterations of one epoch.') if 'mAP' in metric: - xs = np.arange(1, max(epochs) + 1) + xs = np.arange( + int(args.start_epoch), + max(epochs) + 1, int(args.eval_interval)) ys = [] for epoch in epochs: ys += log_dict[epoch][metric] @@ -104,6 +117,16 @@ def add_plot_parser(subparsers): nargs='+', default=['bbox_mAP'], help='the metric that you want to plot') + parser_plt.add_argument( + '--start-epoch', + type=str, + default='1', + help='the epoch that you want to start') + parser_plt.add_argument( + '--eval-interval', + type=str, + default='1', + help='the eval interval when training') parser_plt.add_argument('--title', type=str, help='title of figure') parser_plt.add_argument( '--legend',	./tools/analysis_tools/analyze_logs.py plot_curve IndexError: list index out of range `(openmmlab) lbc@prust-System-3:~/mmdetection-master$ python3.8 ./tools/analysis_tools/analyze_logs.py plot_curve ./work_dirs/deformable_detr_twostage_refine_r50_16x2_50e_coco/20211119_170702.log.json --keys bbox_mAP plot curve of ./work_dirs/deformable_detr_twostage_refine_r50_16x2_50e_coco/20211119_170702.log.json, metric is bbox_mAP Traceback (most recent call last): File "./tools/analysis_tools/analyze_logs.py", line 180, in <module> main() File "./tools/analysis_tools/analyze_logs.py", line 176, in main eval(args.task)(log_dicts, args) File "./tools/analysis_tools/analyze_logs.py", line 53, in plot_curve if metric not in log_dict[epochs[0]]: IndexError: list index out of range `		2022-03-15T10:49:28Z	[]	[]	Traceback (most recent call last): File "./tools/analysis_tools/analyze_logs.py", line 180, in <module> main() File "./tools/analysis_tools/analyze_logs.py", line 176, in main eval(args.task)(log_dicts, args) File "./tools/analysis_tools/analyze_logs.py", line 53, in plot_curve if metric not in log_dict[epochs[0]]: IndexError: list index out of range	10,680
open-mmlab/mmdetection	open-mmlab__mmdetection-8273	ca11860f4f3c3ca2ce8340e2686eeaec05b29111	diff --git a/mmdet/core/hook/wandblogger_hook.py b/mmdet/core/hook/wandblogger_hook.py --- a/mmdet/core/hook/wandblogger_hook.py +++ b/mmdet/core/hook/wandblogger_hook.py @@ -135,7 +135,8 @@ def before_run(self, runner): super(MMDetWandbHook, self).before_run(runner) # Save and Log config. - if runner.meta is not None: + if runner.meta is not None and runner.meta.get('exp_name', + None) is not None: src_cfg_path = osp.join(runner.work_dir, runner.meta.get('exp_name', None)) if osp.exists(src_cfg_path):	WandbLogger Hook Error WandbLogger Hook error Code ``` log_config = dict( interval=10, hooks=[ dict(type='TensorboardLoggerHook'), dict(type='TextLoggerHook'), dict(type='MMDetWandbHook', init_kwargs={ 'project': PROJECT, 'entity': ENTITY, 'name': TAG, 'config': { 'lr': 0.0025, 'batch_size':16 }, 'tags': WANDB_TAGS }, interval=1, log_checkpoint=False, log_checkpoint_metadata=True, num_eval_images=3, bbox_score_thr=0.3 ) ]) ``` Environment ``` Python: 3.7.11 (default, Jul 27 2021, 14:32:16) [GCC 7.5.0] CUDA available: True GPU 0: Tesla V100-SXM2-16GB CUDA_HOME: /usr/local/cuda-10.1 NVCC: Cuda compilation tools, release 10.1, V10.1.24 GCC: gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 PyTorch: 1.10.0+cu102 PyTorch compiling details: PyTorch built with: - GCC 7.3 - C++ Version: 201402 - Intel(R) oneAPI Math Kernel Library Version 2021.4-Product Build 20210904 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.2.3 (Git Hash 7336ca9f055cf1bfa13efb658fe15dc9b41f0740) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70 - CuDNN 7.6.5 - Magma 2.5.2 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=10.2, CUDNN_VERSION=7.6.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.10.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, TorchVision: 0.11.1+cu102 OpenCV: 4.5.4 MMCV: 1.5.3 MMCV Compiler: GCC 7.3 MMCV CUDA Compiler: 10.2 MMDetection: 2.25.0+55d536e ``` Error traceback ``` Traceback (most recent call last): File "train.py", line 28, in <module> fire.Fire(launch) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 141, in Fire component_trace = _Fire(component, args, parsed_flag_args, context, name) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 471, in _Fire target=component.__name__) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 681, in _CallAndUpdateTrace component = fn(varargs, kwargs) File "train.py", line 25, in launch train_detector(model, datasets, cfg, distributed=False, validate=True) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/apis/train.py", line 244, in train_detector runner.run(data_loaders, cfg.workflow) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 111, in run self.call_hook('before_run') File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/base_runner.py", line 309, in call_hook getattr(hook, fn_name)(self) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/dist_utils.py", line 135, in wrapper return func(args, *kwargs) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/core/hook/wandblogger_hook.py", line 140, in before_run runner.meta.get('exp_name', None)) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/posixpath.py", line 94, in join genericpath._check_arg_types('join', a, p) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/genericpath.py", line 153, in _check_arg_types (funcname, s.__class__.__name__)) from None TypeError: join() argument must be str or bytes, not 'NoneType' ```	Hey @sarmientoj24 this seems like a quick fix. ``` # Save and Log config. if runner.meta is not None and runner.meta.get('exp_name', None) is not None: src_cfg_path = osp.join(runner.work_dir, runner.meta.get('exp_name', None)) if osp.exists(src_cfg_path): self.wandb.save(src_cfg_path, base_path=runner.work_dir) self._update_wandb_config(runner) else: runner.logger.warning('No meta information found in the runner. ') ``` I will make a PR to fix this.	2022-06-27T11:30:01Z	[]	[]	Traceback (most recent call last): File "train.py", line 28, in <module> fire.Fire(launch) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 141, in Fire component_trace = _Fire(component, args, parsed_flag_args, context, name) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 471, in _Fire target=component.__name__) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 681, in _CallAndUpdateTrace component = fn(varargs, kwargs) File "train.py", line 25, in launch train_detector(model, datasets, cfg, distributed=False, validate=True) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/apis/train.py", line 244, in train_detector runner.run(data_loaders, cfg.workflow) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 111, in run self.call_hook('before_run') File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/base_runner.py", line 309, in call_hook getattr(hook, fn_name)(self) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/dist_utils.py", line 135, in wrapper return func(args, *kwargs) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/core/hook/wandblogger_hook.py", line 140, in before_run runner.meta.get('exp_name', None)) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/posixpath.py", line 94, in join genericpath._check_arg_types('join', a, p) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/genericpath.py", line 153, in _check_arg_types (funcname, s.__class__.__name__)) from None TypeError: join() argument must be str or bytes, not 'NoneType'	10,692
open-mmlab/mmdetection	open-mmlab__mmdetection-854	ae856e11ec3f281ce77c3f8fdf1cd87162598ffb	diff --git a/mmdet/models/backbones/resnet.py b/mmdet/models/backbones/resnet.py --- a/mmdet/models/backbones/resnet.py +++ b/mmdet/models/backbones/resnet.py @@ -297,11 +297,11 @@ def make_res_layer(block, layers = [] layers.append( block( - inplanes, - planes, - stride, - dilation, - downsample, + inplanes=inplanes, + planes=planes, + stride=stride, + dilation=dilation, + downsample=downsample, style=style, with_cp=with_cp, conv_cfg=conv_cfg, @@ -314,10 +314,10 @@ def make_res_layer(block, for i in range(1, blocks): layers.append( block( - inplanes, - planes, - 1, - dilation, + inplanes=inplanes, + planes=planes, + stride=1, + dilation=dilation, style=style, with_cp=with_cp, conv_cfg=conv_cfg, diff --git a/mmdet/models/backbones/resnext.py b/mmdet/models/backbones/resnext.py --- a/mmdet/models/backbones/resnext.py +++ b/mmdet/models/backbones/resnext.py @@ -11,12 +11,12 @@ class Bottleneck(_Bottleneck): - def __init__(self, groups=1, base_width=4, args, kwargs): + def __init__(self, inplanes, planes, groups=1, base_width=4, kwargs): """Bottleneck block for ResNeXt. If style is "pytorch", the stride-two layer is the 3x3 conv layer, if it is "caffe", the stride-two layer is the first 1x1 conv layer. """ - super(Bottleneck, self).__init__(args, kwargs) + super(Bottleneck, self).__init__(inplanes, planes, kwargs) if groups == 1: width = self.planes	GCNNet with x101 backbone is not work I use gcb in x101 backbone，but get this error: Traceback (most recent call last): File "./tools/train.py", line 95, in <module> main() File "./tools/train.py", line 73, in main cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 60, in build_detector return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/detectors/cascade_rcnn.py", line 36, in __init__ self.backbone = builder.build_backbone(backbone) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 36, in build_backbone return build(cfg, BACKBONES) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 191, in __init__ super(ResNeXt, self).__init__(*kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 447, in __init__ gen_attention_blocks=stage_with_gen_attention[i]) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 317, in make_res_layer (0 in gen_attention_blocks) else None)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 19, in __init__ super(Bottleneck, self).__init__(args, kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 205, in __init__ self.context_block = ContextBlock(inplanes=gcb_inplanes, gcb) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/ops/gcb/context_block.py", line 40, in __init__ nn.Conv2d(self.inplanes, self.planes, kernel_size=1), File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 327, in __init__ False, _pair(0), groups, bias, padding_mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 45, in __init__ self.reset_parameters() File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 48, in reset_parameters init.kaiming_uniform_(self.weight, a=math.sqrt(5)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 323, in kaiming_uniform_ fan = _calculate_correct_fan(tensor, mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 292, in _calculate_correct_fan fan_in, fan_out = _calculate_fan_in_and_fan_out(tensor) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 222, in _calculate_fan_in_and_fan_out receptive_field_size = tensor[0][0].numel() IndexError: index 0 is out of bounds for dimension 0 with size 0	i meet the same problem. Do you know what's wrong? you should update your backbone file，because the orfer of parameters is changed to compliance with specificatinos. See pull #780 Thanks for reporting! Will fix soon. @zhengye1995 i don't know what's your mean? your mean is my hyper parameter have some wrong? but when i run r4,it is ok, i only change gcb 1/4 to 1/16, the problem will appear my backbone as following: backbone=dict( type='ResNeXt', depth=101, groups=32, base_width=4, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, style='pytorch', gcb = dict( ratio=1. / 16., ), stage_with_gcb = (False, True, True, True), dcn=dict( modulated=False, groups=32, deformable_groups=1, fallback_on_stride=False), stage_with_dcn=(False, True, True, True), norm_cfg=norm_cfg, norm_eval=False, ),	2019-06-22T15:25:21Z	[]	[]	Traceback (most recent call last): File "./tools/train.py", line 95, in <module> main() File "./tools/train.py", line 73, in main cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 60, in build_detector return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/detectors/cascade_rcnn.py", line 36, in __init__ self.backbone = builder.build_backbone(backbone) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 36, in build_backbone return build(cfg, BACKBONES) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 191, in __init__ super(ResNeXt, self).__init__(*kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 447, in __init__ gen_attention_blocks=stage_with_gen_attention[i]) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 317, in make_res_layer (0 in gen_attention_blocks) else None)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 19, in __init__ super(Bottleneck, self).__init__(args, kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 205, in __init__ self.context_block = ContextBlock(inplanes=gcb_inplanes, gcb) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/ops/gcb/context_block.py", line 40, in __init__ nn.Conv2d(self.inplanes, self.planes, kernel_size=1), File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 327, in __init__ False, _pair(0), groups, bias, padding_mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 45, in __init__ self.reset_parameters() File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 48, in reset_parameters init.kaiming_uniform_(self.weight, a=math.sqrt(5)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 323, in kaiming_uniform_ fan = _calculate_correct_fan(tensor, mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 292, in _calculate_correct_fan fan_in, fan_out = _calculate_fan_in_and_fan_out(tensor) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 222, in _calculate_fan_in_and_fan_out receptive_field_size = tensor[0][0].numel() IndexError: index 0 is out of bounds for dimension 0 with size 0	10,694
open-mmlab/mmdetection	open-mmlab__mmdetection-9151	e71b499608e9c3ccd4211e7c815fa20eeedf18a2	diff --git a/mmdet/models/detectors/rpn.py b/mmdet/models/detectors/rpn.py --- a/mmdet/models/detectors/rpn.py +++ b/mmdet/models/detectors/rpn.py @@ -1,5 +1,6 @@ # Copyright (c) OpenMMLab. All rights reserved. import warnings +from inspect import signature import mmcv import torch @@ -153,7 +154,9 @@ def show_result(self, data, result, top_k=20, kwargs): np.ndarray: The image with bboxes drawn on it. """ if kwargs is not None: - kwargs.pop('score_thr', None) - kwargs.pop('text_color', None) - kwargs['colors'] = kwargs.pop('bbox_color', 'green') + kwargs['colors'] = 'green' + sig = signature(mmcv.imshow_bboxes) + for k in list(kwargs.keys()): + if k not in sig.parameters: + kwargs.pop(k) mmcv.imshow_bboxes(data, result, top_k=top_k, kwargs)	run image_demo.py with cascade_rpn model show error ### Prerequisite - [X] I have searched [the existing and past issues](https://github.com/open-mmlab/mmdetection/issues) but cannot get the expected help. - [X] I have read the [FAQ documentation](https://mmdetection.readthedocs.io/en/latest/faq.html) but cannot get the expected help. - [X] The bug has not been fixed in the [latest version](https://github.com/open-mmlab/mmdetection). ### 🐞 Describe the bug ``` python demo/image_demo.py demo/demo.jpg configs/cascade_rpn/crpn_r50_caffe_fpn_1x_coco.py checkpoints/cascade_rpn_r50_caffe_fpn_1x_coco-7aa93cef.pth``` ``` UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:2895.) return _VF.meshgrid(tensors, kwargs) # type: ignore[attr-defined] Traceback (most recent call last): File "demo/image_demo.py", line 68, in <module> main(args) File "demo/image_demo.py", line 38, in main show_result_pyplot( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/apis/inference.py", line 241, in show_result_pyplot model.show_result( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/models/detectors/rpn.py", line 159, in show_result mmcv.imshow_bboxes(data, result, top_k=top_k, kwargs) TypeError: imshow_bboxes() got an unexpected keyword argument 'mask_color' ``` ### Environment ``` Python: 3.8.13 (default, Mar 28 2022, 11:38:47) [GCC 7.5.0] CUDA available: True GPU 0: GeForce RTX 2080 SUPER CUDA_HOME: /usr/local/cuda-10.2 NVCC: Cuda compilation tools, release 10.2, V10.2.8 GCC: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609 PyTorch: 1.12.1+cu102 PyTorch compiling details: PyTorch built with: - GCC 7.3 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70 - CuDNN 7.6.5 - Magma 2.5.2 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=10.2, CUDNN_VERSION=7.6.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.12.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=OFF, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, TorchVision: 0.13.1+cu102 OpenCV: 4.6.0 MMCV: 1.6.2 MMCV Compiler: GCC 7.3 MMCV CUDA Compiler: 10.2 MMDetection: 2.25.2+9d3e162 ``` ### Additional information _No response_	Thanks for your report! We will fix this bug in the next version.	2022-10-27T03:49:51Z	[]	[]	Traceback (most recent call last): File "demo/image_demo.py", line 68, in <module> main(args) File "demo/image_demo.py", line 38, in main show_result_pyplot( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/apis/inference.py", line 241, in show_result_pyplot model.show_result( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/models/detectors/rpn.py", line 159, in show_result mmcv.imshow_bboxes(data, result, top_k=top_k, **kwargs) TypeError: imshow_bboxes() got an unexpected keyword argument 'mask_color'	10,695
open-mmlab/mmdetection	open-mmlab__mmdetection-9694	ea29a1edbe973389c2d705d99fd060b07decddab	diff --git a/mmdet/utils/__init__.py b/mmdet/utils/__init__.py --- a/mmdet/utils/__init__.py +++ b/mmdet/utils/__init__.py @@ -7,6 +7,7 @@ from .memory import AvoidCUDAOOM, AvoidOOM from .misc import find_latest_checkpoint, update_data_root from .replace_cfg_vals import replace_cfg_vals +from .rfnext import rfnext_init_model from .setup_env import setup_multi_processes from .split_batch import split_batch from .util_distribution import build_ddp, build_dp, get_device @@ -16,5 +17,6 @@ 'update_data_root', 'setup_multi_processes', 'get_caller_name', 'log_img_scale', 'compat_cfg', 'split_batch', 'build_ddp', 'build_dp', 'get_device', 'replace_cfg_vals', 'AvoidOOM', 'AvoidCUDAOOM', - 'get_max_num_gt_division_factor', 'masked_fill', 'batch_images_to_levels' + 'get_max_num_gt_division_factor', 'masked_fill', 'batch_images_to_levels', + 'rfnext_init_model' ]	[Bug] Training Error ### Prerequisite - [X] I have searched [Issues](https://github.com/open-mmlab/mmdetection/issues) and [Discussions](https://github.com/open-mmlab/mmdetection/discussions) but cannot get the expected help. - [X] I have read the [FAQ documentation](https://mmdetection.readthedocs.io/en/latest/faq.html) but cannot get the expected help. - [X] The bug has not been fixed in the [latest version (master)](https://github.com/open-mmlab/mmdetection) or [latest version (3.x)](https://github.com/open-mmlab/mmdetection/tree/dev-3.x). ### Task I have modified the scripts/configs, or I'm working on my own tasks/models/datasets. ### Branch master branch https://github.com/open-mmlab/mmdetection ### Environment ``` sys.platform: linux Python: 3.7.15 (default, Nov 24 2022, 21:12:53) [GCC 11.2.0] CUDA available: True GPU 0,1,2,3,4,5,6,7: NVIDIA GeForce RTX 3090 CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 11.6, V11.6.55 GCC: gcc (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0 PyTorch: 1.13.1+cu117 PyTorch compiling details: PyTorch built with: - GCC 9.3 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 11.7 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86 - CuDNN 8.5 - Magma 2.6.1 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.13.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, TorchVision: 0.14.1+cu117 OpenCV: 4.7.0 MMCV: 1.7.1 MMCV Compiler: GCC 9.3 MMCV CUDA Compiler: 11.7 MMDetection: 2.28.0+b955832 ``` ### Reproduces the problem - code sample ```python python train.py $CONFIG_FILE ``` ### Reproduces the problem - command or script ```python python train.py $CONFIG_FILE ``` ### Reproduces the problem - error message ``` /anaconda3/envs/dl/lib/python3.7/site-packages/mmcv/__init__.py:21: UserWarning: On January 1, 2023, MMCV will release v2.0.0, in which it will remove components related to the training process and add a data transformation module. In addition, it will rename the package names mmcv to mmcv-lite and mmcv-full to mmcv. See https://github.com/open-mmlab/mmcv/blob/master/docs/en/compatibility.md for more details. 'On January 1, 2023, MMCV will release v2.0.0, in which it will remove ' Traceback (most recent call last): File "train.py", line 20, in <module> from mmdet.utils import (collect_env, get_device, get_root_logger, ImportError: cannot import name 'rfnext_init_model' from 'mmdet.utils' (/mmdetection/mmdet/utils/__init__.py) ``` ### Additional information I clone the latest mmdetection package and intend to run a couple of benchmark. However, neither my configs nor the configs provided in the `mmdet/configs` yields the abovementioned error message.	I also encountered this problem	2023-01-30T08:09:21Z	[]	[]	Traceback (most recent call last): File "train.py", line 20, in <module> from mmdet.utils import (collect_env, get_device, get_root_logger, ImportError: cannot import name 'rfnext_init_model' from 'mmdet.utils' (/mmdetection/mmdet/utils/__init__.py)	10,701
pandas-dev/pandas	pandas-dev__pandas-10108	eafd22d961934a7b3cc72607ef4512a18b419085	diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -57,3 +57,5 @@ Performance Improvements Bug Fixes ~~~~~~~~~ + +- Bug in ``Categorical`` repr with ``display.width`` of ``None`` in Python 3 (:issue:`10087`) diff --git a/pandas/core/categorical.py b/pandas/core/categorical.py --- a/pandas/core/categorical.py +++ b/pandas/core/categorical.py @@ -1310,8 +1310,7 @@ def _repr_categories_info(self): levheader = "Categories (%d, %s): " % (len(self.categories), self.categories.dtype) width, height = get_terminal_size() - max_width = (width if get_option("display.width") == 0 - else get_option("display.width")) + max_width = get_option("display.width") or width if com.in_ipython_frontend(): # 0 = no breaks max_width = 0	BUG: categorical doesn't handle display.width of None in Python 3 Categorical Series have a special repr that looks at display.width, which can be None if following the Options and Settings docs. Unlike Python 2, in Python 3 an integer vs None comparison throws an exception. (on current master, and has been true for several releases now) ``` python Python 3.4.0 (default, Apr 11 2014, 13:05:11) [GCC 4.8.2] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import pandas as pd >>> pd.core.config.set_option('display.width', None) >>> import numpy as np >>> x = pd.Series(np.random.randn(100)) >>> pd.cut(x, 10) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 67, in __repr__ return str(self) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 46, in __str__ return self.__unicode__() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 897, in __unicode__ max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 962, in to_string name=name, max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 992, in _get_repr result = formatter.to_string() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 222, in to_string footer = self._get_footer() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 196, in _get_footer level_info = self.tr_series.values._repr_categories_info() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/categorical.py", line 1323, in _repr_categories_info if max_width != 0 and cur_col_len + sep_len + len(val) > max_width: TypeError: unorderable types: int() > NoneType() ```	should be a simple fix. PR's are welcome. Sure, I'll try and get to it soon.	2015-05-11T22:15:45Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 67, in __repr__ return str(self) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 46, in __str__ return self.__unicode__() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 897, in __unicode__ max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 962, in to_string name=name, max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 992, in _get_repr result = formatter.to_string() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 222, in to_string footer = self._get_footer() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 196, in _get_footer level_info = self.tr_series.values._repr_categories_info() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/categorical.py", line 1323, in _repr_categories_info if max_width != 0 and cur_col_len + sep_len + len(val) > max_width: TypeError: unorderable types: int() > NoneType()	10,720
pandas-dev/pandas	pandas-dev__pandas-10473	83b232089b9292b11d4b9b00c0e50cc4a829f016	diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -27,6 +27,7 @@ New features ~~~~~~~~~~~~ - SQL io functions now accept a SQLAlchemy connectable. (:issue:`7877`) +- Enable writing complex values to HDF stores when using table format (:issue:`10447`) .. _whatsnew_0170.enhancements.other: @@ -147,3 +148,4 @@ Bug Fixes - Bug in `groupby.var` which caused variance to be inaccurate for small float values (:issue:`10448`) - Bug in ``Series.plot(kind='hist')`` Y Label not informative (:issue:`10485`) + diff --git a/pandas/io/pytables.py b/pandas/io/pytables.py --- a/pandas/io/pytables.py +++ b/pandas/io/pytables.py @@ -1773,6 +1773,8 @@ def set_kind(self): self.kind = 'string' elif dtype.startswith(u('float')): self.kind = 'float' + elif dtype.startswith(u('complex')): + self.kind = 'complex' elif dtype.startswith(u('int')) or dtype.startswith(u('uint')): self.kind = 'integer' elif dtype.startswith(u('date')): @@ -1802,6 +1804,8 @@ def set_atom(self, block, block_items, existing_col, min_itemsize, return self.set_atom_datetime64(block) elif block.is_timedelta: return self.set_atom_timedelta64(block) + elif block.is_complex: + return self.set_atom_complex(block) dtype = block.dtype.name inferred_type = lib.infer_dtype(block.values) @@ -1936,6 +1940,12 @@ def get_atom_coltype(self, kind=None): def get_atom_data(self, block, kind=None): return self.get_atom_coltype(kind=kind)(shape=block.shape[0]) + def set_atom_complex(self, block): + self.kind = block.dtype.name + itemsize = int(self.kind.split('complex')[-1]) // 8 + self.typ = _tables().ComplexCol(itemsize=itemsize, shape=block.shape[0]) + self.set_data(block.values.astype(self.typ.type, copy=False)) + def set_atom_data(self, block): self.kind = block.dtype.name self.typ = self.get_atom_data(block) @@ -3147,8 +3157,8 @@ def f(i, c): def create_index(self, columns=None, optlevel=None, kind=None): """ Create a pytables index on the specified columns - note: cannot index Time64Col() currently; PyTables must be >= 2.3 - + note: cannot index Time64Col() or ComplexCol currently; + PyTables must be >= 3.0 Paramaters ---------- @@ -3203,6 +3213,12 @@ def create_index(self, columns=None, optlevel=None, kind=None): # create the index if not v.is_indexed: + if v.type.startswith('complex'): + raise TypeError('Columns containing complex values can be stored but cannot' + ' be indexed when using table format. Either use fixed ' + 'format, set index=False, or do not include the columns ' + 'containing complex values to data_columns when ' + 'initializing the table.') v.create_index(kw) def read_axes(self, where, kwargs):	BUG: Cannot store complex valued Series/DataFrame/Panel/Panel4D as 'table' to hdf Attempting to store a Panel4D of complext128 results in an error. This is true if I use table or fixed formats (different errors) [I am showing the last call from pandas as well as the actual error] ``` python pd.Panel4D(np.tile(np.array([1+1j,2+2j]),[2,2,2,1])).to_hdf('complex.h5','complex',format='t') Traceback (most recent call last): ... File "/miniconda/envs/py34/lib/python3.4/site-packages/pandas/io/pytables.py", line 1930, in get_atom_data return self.get_atom_coltype(kind=kind)(shape=block.shape[0]) ... File "/miniconda/envs/py34/lib/python3.4/site-packages/tables/atom.py", line 740, in __init__ "to avoid confusions with PyTables 1.X complex atom names, " TypeError: to avoid confusions with PyTables 1.X complex atom names, please use ``ComplexAtom(itemsize=N)``, where N=8 for single precision complex atoms, and N=16 for double precision complex atoms ``` ``` INSTALLED VERSIONS ------------------ commit: None python: 3.4.3.final.0 python-bits: 64 OS: Linux OS-release: 2.6.32-504.16.2.el6.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 pandas: 0.16.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: None statsmodels: None IPython: 3.1.0 sphinx: None patsy: 0.3.0 dateutil: 2.4.2 pytz: 2015.4 bottleneck: None tables: 3.1.1 numexpr: 2.3.1 matplotlib: None openpyxl: 2.0.2 xlrd: 0.9.3 xlwt: 1.0.0 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None ``` If you try fixed, you get ``` TypeError: cannot properly create the storer for: [_STORER_MAP] [group->/complex (Group) '',value-><class 'pandas.core.panelnd.Panel4D'>,format->fixed,append->False,kwargs->{'encoding': None}] ```	'>3d' not supported in fixed at all (as pretty useless - u generally need/want to append to these) complex is not supported very well in the exporters but of course fixes welcome Turns out this is true for all types, not just Panel4D	2015-06-29T19:27:56Z	[]	[]	Traceback (most recent call last): ... File "/miniconda/envs/py34/lib/python3.4/site-packages/pandas/io/pytables.py", line 1930, in get_atom_data return self.get_atom_coltype(kind=kind)(shape=block.shape[0]) ... File "/miniconda/envs/py34/lib/python3.4/site-packages/tables/atom.py", line 740, in __init__ "to avoid confusions with PyTables 1.X complex atom names, " TypeError: to avoid confusions with PyTables 1.X complex atom names, please use ``ComplexAtom(itemsize=N)``, where N=8 for single precision complex atoms, and N=16 for double precision complex atoms	10,768
pandas-dev/pandas	pandas-dev__pandas-10497	bbec57d6f881cb7d26ad65319595c8594381fe8c	diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -128,6 +128,7 @@ Bug Fixes - Bug in ``test_categorical`` on big-endian builds (:issue:`10425`) +- Bug in ``Series.shift`` and ``DataFrame.shift`` not supporting categorical data (:issue:`9416`) - Bug in ``Series.map`` using categorical ``Series`` raises ``AttributeError`` (:issue:`10324`) - Bug in ``MultiIndex.get_level_values`` including ``Categorical`` raises ``AttributeError`` (:issue:`10460`) diff --git a/pandas/core/categorical.py b/pandas/core/categorical.py --- a/pandas/core/categorical.py +++ b/pandas/core/categorical.py @@ -820,6 +820,35 @@ def shape(self): return tuple([len(self._codes)]) + def shift(self, periods): + """ + Shift Categorical by desired number of periods. + + Parameters + ---------- + periods : int + Number of periods to move, can be positive or negative + + Returns + ------- + shifted : Categorical + """ + # since categoricals always have ndim == 1, an axis parameter + # doesnt make any sense here. + codes = self.codes + if codes.ndim > 1: + raise NotImplementedError("Categorical with ndim > 1.") + if np.prod(codes.shape) and (periods != 0): + codes = np.roll(codes, com._ensure_platform_int(periods), axis=0) + if periods > 0: + codes[:periods] = -1 + else: + codes[periods:] = -1 + + return Categorical.from_codes(codes, + categories=self.categories, + ordered=self.ordered) + def __array__(self, dtype=None): """ The numpy array interface. diff --git a/pandas/core/internals.py b/pandas/core/internals.py --- a/pandas/core/internals.py +++ b/pandas/core/internals.py @@ -1709,6 +1709,10 @@ def interpolate(self, method='pad', axis=0, inplace=False, limit=limit), placement=self.mgr_locs) + def shift(self, periods, axis=0): + return self.make_block_same_class(values=self.values.shift(periods), + placement=self.mgr_locs) + def take_nd(self, indexer, axis=0, new_mgr_locs=None, fill_tuple=None): """ Take values according to indexer and return them as a block.bb	Series.shift() doesn't work for categorical type Not sure if this is intentional, but Series.shift() won't run with categorical dtypes: ``` python ser = pd.Series(['a', 'b', 'c', 'd'], dtype="category") ser.shift(1) Traceback (most recent call last): File "<ipython-input-15-1a7536b0af06>", line 1, in <module> ser.shift(1) File "/.../pandas/core/generic.py", line 3394, in shift new_data = self._data.shift(periods=periods, axis=block_axis) File "/.../pandas/core/internals.py", line 2533, in shift return self.apply('shift', kwargs) File "/.../pandas/core/internals.py", line 2497, in apply applied = getattr(b, f)(kwargs) File "/.../pandas/core/internals.py", line 893, in shift new_values, fill_value = com._maybe_upcast(self.values) File "/.../pandas/core/common.py", line 1218, in _maybe_upcast new_dtype, fill_value = _maybe_promote(dtype, fill_value) File "/.../pandas/core/common.py", line 1124, in _maybe_promote if issubclass(np.dtype(dtype).type, compat.string_types): TypeError: data type not understood ```	This simply hasn't been implemented, but otherwise was not intentional. Help would be appreciated if you're interested in putting together a PR. The place to get started (I believe) would be to implement the `shift` method on `CategoricalBlock` in `pandas.core.internals`. here's basically what you would do ``` In [2]: s = Series(list('aabbcde'),dtype='category') In [3]: s Out[3]: 0 a 1 a 2 b 3 b 4 c 5 d 6 e dtype: category Categories (5, object): [a < b < c < d < e] In [4]: s.values Out[4]: [a, a, b, b, c, d, e] Categories (5, object): [a < b < c < d < e] In [5]: s.values.codes Out[5]: array([0, 0, 1, 1, 2, 3, 4], dtype=int8) In [6]: np.roll(s.values.codes,len(s)-1,axis=0) Out[6]: array([0, 1, 1, 2, 3, 4, 0], dtype=int8) In [7]: codes = np.roll(s.values.codes,len(s)-1,axis=0) In [8]: codes[-1] = -1 In [11]: pd.Categorical(codes,categories=s.values.categories,fastpath=True) Out[11]: [a, b, b, c, d, e, NaN] Categories (5, object): [a, b, c, d, e] ``` you would use the Block.shift method (and pass the codes to it for the actual shifting), then wrap it back to a catetgorical (their is a method for that too). Should be pretty straightforward.	2015-07-03T13:42:01Z	[]	[]	Traceback (most recent call last): File "<ipython-input-15-1a7536b0af06>", line 1, in <module> ser.shift(1) File "/.../pandas/core/generic.py", line 3394, in shift new_data = self._data.shift(periods=periods, axis=block_axis) File "/.../pandas/core/internals.py", line 2533, in shift return self.apply('shift', kwargs) File "/.../pandas/core/internals.py", line 2497, in apply applied = getattr(b, f)(kwargs) File "/.../pandas/core/internals.py", line 893, in shift new_values, fill_value = com._maybe_upcast(self.values) File "/.../pandas/core/common.py", line 1218, in _maybe_upcast new_dtype, fill_value = _maybe_promote(dtype, fill_value) File "/.../pandas/core/common.py", line 1124, in _maybe_promote if issubclass(np.dtype(dtype).type, compat.string_types): TypeError: data type not understood	10,770
pandas-dev/pandas	pandas-dev__pandas-10853	d27068f8b78661a64580340a5ab230d0dad17760		testArrayNumpyLabelled fails on Python 2.7.10 ``` FAIL: testArrayNumpyLabelled (pandas.io.tests.test_json.test_ujson.NumpyJSONTests) ---------------------------------------------------------------------- Traceback (most recent call last): File "/Users/ch/repo/pandas/.tox/py27/lib/python2.7/site-packages/pandas/io/tests/test_json/test_ujson.py", line 1141, in testArrayNumpyLabelled self.assertTrue((expectedvals == output[0]).all()) AssertionError: False is not true -------------------- >> begin captured stdout << --------------------- [[42 31] [24 99] [ 2 78]] [[31 42] [99 24] [78 2]] --------------------- >> end captured stdout << ---------------------- ``` Order of dict elements seems to be non-deterministic for python2.7 versions as well.	well this doesn't fail on travis AFAICT, nor have I ever actually seen this fail. So not sure how to repro. Do you have a 32-bit build? This test does seem to depend on a few things, including architecture and the PYTHONHASHSEED environment variable. @kawochen, No, I am running a 64bit interpreter. Yes, the failure is highly hardware/interpreter/seed dependent. But that is the point of this bug report. @cel4 can you give the specific system this fails on. Yes its is somewhat non-deterministic, though in py2 the ordering is deterministic. Kind of a silly test actually (but IIRC came from another package :). Why don't you do a pull-request to sort them first. ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Darwin OS-release: 14.5.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: de_DE.UTF-8 LANG: de_DE.UTF-8 pandas: 0.16.2+340.ge4368de.dirty nose: 1.3.7 Cython: 0.23 numpy: 1.9.2 scipy: None statsmodels: None IPython: None sphinx: None patsy: None dateutil: 2.4.2 pytz: 2015.4 bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: 1.8.6 xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.4.0 html5lib: 0.999999 httplib2: 0.9.1 apiclient: 1.2 sqlalchemy: 1.0.8 pymysql: None psycopg2: None ``` If the test fails or passes depends on the python hashseed. e.g. `PYTHONHASHSEED='1145934480'` deterministically fails that test, `PYTHONHASHSEED='2585294338'` deterministically passes the test here.	2015-08-19T03:59:44Z	[]	[]	Traceback (most recent call last): File "/Users/ch/repo/pandas/.tox/py27/lib/python2.7/site-packages/pandas/io/tests/test_json/test_ujson.py", line 1141, in testArrayNumpyLabelled self.assertTrue((expectedvals == output[0]).all()) AssertionError: False is not true	10,813
pandas-dev/pandas	pandas-dev__pandas-11114	d1feb49267da6074603c6a9dbf6314681378cd81	diff --git a/.travis.yml b/.travis.yml --- a/.travis.yml +++ b/.travis.yml @@ -43,13 +43,6 @@ matrix: - CLIPBOARD_GUI=gtk2 - BUILD_TYPE=conda - DOC_BUILD=true # if rst files were changed, build docs in parallel with tests - - python: 3.3 - env: - - JOB_NAME: "33_nslow" - - NOSE_ARGS="not slow and not disabled" - - FULL_DEPS=true - - CLIPBOARD=xsel - - BUILD_TYPE=conda - python: 3.4 env: - JOB_NAME: "34_nslow" @@ -64,6 +57,13 @@ matrix: - FULL_DEPS=true - CLIPBOARD=xsel - BUILD_TYPE=conda + - python: 3.3 + env: + - JOB_NAME: "33_nslow" + - NOSE_ARGS="not slow and not disabled" + - FULL_DEPS=true + - CLIPBOARD=xsel + - BUILD_TYPE=conda - python: 2.7 env: - JOB_NAME: "27_slow" @@ -104,10 +104,10 @@ matrix: - BUILD_TYPE=pydata - PANDAS_TESTING_MODE="deprecate" allow_failures: - - python: 3.5 + - python: 3.3 env: - - JOB_NAME: "35_nslow" - - NOSE_ARGS="not slow and not network and not disabled" + - JOB_NAME: "33_nslow" + - NOSE_ARGS="not slow and not disabled" - FULL_DEPS=true - CLIPBOARD=xsel - BUILD_TYPE=conda diff --git a/ci/requirements-3.5.txt b/ci/requirements-3.5.txt --- a/ci/requirements-3.5.txt +++ b/ci/requirements-3.5.txt @@ -10,3 +10,15 @@ cython scipy numexpr pytables +html5lib +lxml + +# currently causing some warnings +#sqlalchemy +#pymysql +#psycopg2 + +# not available from conda +#beautiful-soup +#bottleneck +#matplotlib diff --git a/doc/source/install.rst b/doc/source/install.rst --- a/doc/source/install.rst +++ b/doc/source/install.rst @@ -18,7 +18,7 @@ Instructions for installing from source, Python version support ---------------------- -Officially Python 2.6, 2.7, 3.3, and 3.4. +Officially Python 2.6, 2.7, 3.3, 3.4, and 3.5 Installing pandas ----------------- diff --git a/doc/source/release.rst b/doc/source/release.rst --- a/doc/source/release.rst +++ b/doc/source/release.rst @@ -64,6 +64,7 @@ Highlights include: - Support for reading SAS xport files, see :ref:`here <whatsnew_0170.enhancements.sas_xport>` - Documentation comparing SAS to pandas, see :ref:`here <compare_with_sas>` - Removal of the automatic TimeSeries broadcasting, deprecated since 0.8.0, see :ref:`here <whatsnew_0170.prior_deprecations>` +- Compatibility with Python 3.5 See the :ref:`v0.17.0 Whatsnew <whatsnew_0170>` overview for an extensive list of all enhancements and bugs that have been fixed in 0.17.0. diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -49,6 +49,7 @@ Highlights include: - Support for reading SAS xport files, see :ref:`here <whatsnew_0170.enhancements.sas_xport>` - Documentation comparing SAS to pandas, see :ref:`here <compare_with_sas>` - Removal of the automatic TimeSeries broadcasting, deprecated since 0.8.0, see :ref:`here <whatsnew_0170.prior_deprecations>` +- Compatibility with Python 3.5 (:issue:`11097`) Check the :ref:`API Changes <whatsnew_0170.api>` and :ref:`deprecations <whatsnew_0170.deprecations>` before updating. diff --git a/pandas/compat/__init__.py b/pandas/compat/__init__.py --- a/pandas/compat/__init__.py +++ b/pandas/compat/__init__.py @@ -36,9 +36,9 @@ import sys import types -PY3 = (sys.version_info[0] >= 3) PY2 = sys.version_info[0] == 2 - +PY3 = (sys.version_info[0] >= 3) +PY35 = (sys.version_info >= (3, 5)) try: import __builtin__ as builtins diff --git a/pandas/computation/expr.py b/pandas/computation/expr.py --- a/pandas/computation/expr.py +++ b/pandas/computation/expr.py @@ -516,7 +516,54 @@ def visit_Attribute(self, node, kwargs): raise ValueError("Invalid Attribute context {0}".format(ctx.__name__)) - def visit_Call(self, node, side=None, kwargs): + def visit_Call_35(self, node, side=None, *kwargs): + """ in 3.5 the starargs attribute was changed to be more flexible, #11097 """ + + if isinstance(node.func, ast.Attribute): + res = self.visit_Attribute(node.func) + elif not isinstance(node.func, ast.Name): + raise TypeError("Only named functions are supported") + else: + try: + res = self.visit(node.func) + except UndefinedVariableError: + # Check if this is a supported function name + try: + res = FuncNode(node.func.id) + except ValueError: + # Raise original error + raise + + if res is None: + raise ValueError("Invalid function call {0}".format(node.func.id)) + if hasattr(res, 'value'): + res = res.value + + if isinstance(res, FuncNode): + + new_args = [ self.visit(arg) for arg in node.args ] + + if node.keywords: + raise TypeError("Function \"{0}\" does not support keyword " + "arguments".format(res.name)) + + return res(new_args, *kwargs) + + else: + + new_args = [ self.visit(arg).value for arg in node.args ] + + for key in node.keywords: + if not isinstance(key, ast.keyword): + raise ValueError("keyword error in function call " + "'{0}'".format(node.func.id)) + + if key.arg: + kwargs.append(ast.keyword(keyword.arg, self.visit(keyword.value))) + + return self.const_type(res(new_args, kwargs), self.env) + + def visit_Call_legacy(self, node, side=None, kwargs): # this can happen with: datetime.datetime if isinstance(node.func, ast.Attribute): @@ -607,6 +654,13 @@ def visitor(x, y): operands = node.values return reduce(visitor, operands) +# ast.Call signature changed on 3.5, +# conditionally change which methods is named +# visit_Call depending on Python version, #11097 +if compat.PY35: + BaseExprVisitor.visit_Call = BaseExprVisitor.visit_Call_35 +else: + BaseExprVisitor.visit_Call = BaseExprVisitor.visit_Call_legacy _python_not_supported = frozenset(['Dict', 'BoolOp', 'In', 'NotIn']) _numexpr_supported_calls = frozenset(_reductions + _mathops) diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -181,6 +181,7 @@ def build_extensions(self): 'Programming Language :: Python :: 2.7', 'Programming Language :: Python :: 3.3', 'Programming Language :: Python :: 3.4', + 'Programming Language :: Python :: 3.5', 'Programming Language :: Cython', 'Topic :: Scientific/Engineering', ]	TST/COMPAT: python 3.5 compat A few changes are needed: https://travis-ci.org/jreback/pandas/jobs/80331174 (since merged to master) so the Call node as changed need to do something like this: https://bitbucket.org/pytest-dev/pytest/pull-requests/296/astcall-signature-changed-on-35/diff If this is fixed I think everything will pass. I guess this is an API change in python. ``` ====================================================================== ERROR: test_df_use_case (pandas.computation.tests.test_eval.TestMathNumExprPandas) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/travis/build/jreback/pandas/pandas/computation/tests/test_eval.py", line 1488, in test_df_use_case parser=self.parser) File "/home/travis/build/jreback/pandas/pandas/core/frame.py", line 2102, in eval return _eval(expr, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 656, in __init__ self.terms = self.parse() File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 673, in parse return self._visitor.visit(self.expr) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 498, in visit_Assign return self.visit(node.value, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in visit_Call args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in <listcomp> args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 545, in visit_Call if node.starargs is not None: AttributeError: 'Call' object has no attribute 'starargs' ```		2015-09-15T16:32:37Z	[]	[]	Traceback (most recent call last): File "/home/travis/build/jreback/pandas/pandas/computation/tests/test_eval.py", line 1488, in test_df_use_case parser=self.parser) File "/home/travis/build/jreback/pandas/pandas/core/frame.py", line 2102, in eval return _eval(expr, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 656, in __init__ self.terms = self.parse() File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 673, in parse return self._visitor.visit(self.expr) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 498, in visit_Assign return self.visit(node.value, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in visit_Call args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in <listcomp> args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 545, in visit_Call if node.starargs is not None: AttributeError: 'Call' object has no attribute 'starargs'	10,840
pandas-dev/pandas	pandas-dev__pandas-11366	b09b84e8e0baf89e78b618cdda30af11087d2e4a	diff --git a/doc/source/whatsnew/v0.17.1.txt b/doc/source/whatsnew/v0.17.1.txt --- a/doc/source/whatsnew/v0.17.1.txt +++ b/doc/source/whatsnew/v0.17.1.txt @@ -94,7 +94,7 @@ Bug Fixes - +- Bug in ``pd.eval`` where unary ops in a list error (:issue:`11235`) - Bug in ``squeeze()`` with zero length arrays (:issue:`11230`, :issue:`8999`) diff --git a/pandas/computation/expr.py b/pandas/computation/expr.py --- a/pandas/computation/expr.py +++ b/pandas/computation/expr.py @@ -427,7 +427,7 @@ def visit_Str(self, node, kwargs): return self.term_type(name, self.env) def visit_List(self, node, kwargs): - name = self.env.add_tmp([self.visit(e).value for e in node.elts]) + name = self.env.add_tmp([self.visit(e)(self.env) for e in node.elts]) return self.term_type(name, self.env) visit_Tuple = visit_List @@ -655,7 +655,7 @@ def visitor(x, y): return reduce(visitor, operands) # ast.Call signature changed on 3.5, -# conditionally change which methods is named +# conditionally change which methods is named # visit_Call depending on Python version, #11097 if compat.PY35: BaseExprVisitor.visit_Call = BaseExprVisitor.visit_Call_35	Dataframe.eval(): Negative number in list passed to 'in'-expression causes crash on python 3.4.0 The following crashes on python 3.4.0. It works fine on Python 2.7.5. ``` >>> import pandas >>> from io import StringIO >>> data = "foo,bar\n11,12" >>> df = pandas.read_csv(StringIO(data)) >>> df.eval('foo in [11, 32]') 0 True dtype: bool >>> df.eval('foo in [11, -32]') Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/core/frame.py", line 1987, in eval return _eval(expr, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 635, in __init__ self.terms = self.parse() File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 652, in parse return self._visitor.visit(self.expr) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 323, in visit_Expr return self.visit(node.value, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 560, in visit_Compare return self.visit(binop) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 404, in visit_BinOp op, op_class, left, right = self._possibly_transform_eq_ne(node) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 357, in _possibly_transform_eq_ne right = self.visit(node.right, side='right') File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in visit_List name = self.env.add_tmp([self.visit(e).value for e in node.elts]) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in <listcomp> name = self.env.add_tmp([self.visit(e).value for e in node.elts]) AttributeError: 'UnaryOp' object has no attribute 'value' >>> ```	pls show a reproducible example, e.g. show how the actual frame is created so it can be copy-pasted pls show `pd.show_versions()` Sorry, my bad. The lines actually defining the data fell away in during copy-paste. I've updated the example above. Here's show_versions() ``` >>> import pandas >>> pandas.show_versions() INSTALLED VERSIONS ------------------ commit: None python: 3.4.0.final.0 python-bits: 64 OS: Linux OS-release: 3.13.0-48-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 pandas: 0.16.2 nose: None Cython: None numpy: 1.9.3 scipy: None statsmodels: None IPython: None sphinx: None patsy: None dateutil: 2.4.2 pytz: 2015.6 bottleneck: None tables: None numexpr: 2.4.4 matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None ``` yeh I guess this is parsed differently in 3.4 that in 2.7. pull-requests to fix are welcome	2015-10-19T05:40:18Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/core/frame.py", line 1987, in eval return _eval(expr, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 635, in __init__ self.terms = self.parse() File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 652, in parse return self._visitor.visit(self.expr) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 323, in visit_Expr return self.visit(node.value, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 560, in visit_Compare return self.visit(binop) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 404, in visit_BinOp op, op_class, left, right = self._possibly_transform_eq_ne(node) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 357, in _possibly_transform_eq_ne right = self.visit(node.right, side='right') File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in visit_List name = self.env.add_tmp([self.visit(e).value for e in node.elts]) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in <listcomp> name = self.env.add_tmp([self.visit(e).value for e in node.elts]) AttributeError: 'UnaryOp' object has no attribute 'value'	10,873
pandas-dev/pandas	pandas-dev__pandas-11400	88e8d6e7dfcea3435d7695a0c312690a57c05663	diff --git a/doc/source/whatsnew/v0.17.1.txt b/doc/source/whatsnew/v0.17.1.txt --- a/doc/source/whatsnew/v0.17.1.txt +++ b/doc/source/whatsnew/v0.17.1.txt @@ -70,7 +70,7 @@ Bug Fixes - Bug in ``HDFStore.append`` with strings whose encoded length exceded the max unencoded length (:issue:`11234`) - Bug in merging ``datetime64[ns, tz]`` dtypes (:issue:`11405`) - Bug in ``HDFStore.select`` when comparing with a numpy scalar in a where clause (:issue:`11283`) - +- Bug in using ``DataFrame.ix`` with a multi-index indexer(:issue:`11372`) - Bug in tz-conversions with an ambiguous time and ``.dt`` accessors (:issue:`11295`) diff --git a/pandas/core/indexing.py b/pandas/core/indexing.py --- a/pandas/core/indexing.py +++ b/pandas/core/indexing.py @@ -443,11 +443,14 @@ def can_do_equal_len(): # we have an equal len Frame if isinstance(value, ABCDataFrame) and value.ndim > 1: sub_indexer = list(indexer) + multiindex_indexer = isinstance(labels, MultiIndex) for item in labels: if item in value: sub_indexer[info_axis] = item - v = self._align_series(tuple(sub_indexer), value[item]) + v = self._align_series( + tuple(sub_indexer), value[item], multiindex_indexer + ) else: v = np.nan @@ -516,8 +519,28 @@ def can_do_equal_len(): self.obj._data = self.obj._data.setitem(indexer=indexer, value=value) self.obj._maybe_update_cacher(clear=True) - def _align_series(self, indexer, ser): - # indexer to assign Series can be tuple, slice, scalar + def _align_series(self, indexer, ser, multiindex_indexer=False): + """ + Parameters + ---------- + indexer : tuple, slice, scalar + The indexer used to get the locations that will be set to + `ser` + + ser : pd.Series + The values to assign to the locations specified by `indexer` + + multiindex_indexer : boolean, optional + Defaults to False. Should be set to True if `indexer` was from + a `pd.MultiIndex`, to avoid unnecessary broadcasting. + + + Returns: + -------- + `np.array` of `ser` broadcast to the appropriate shape for assignment + to the locations selected by `indexer` + + """ if isinstance(indexer, (slice, np.ndarray, list, Index)): indexer = tuple([indexer]) @@ -555,7 +578,7 @@ def _align_series(self, indexer, ser): ser = ser.reindex(obj.axes[0][indexer[0]], copy=True)._values # single indexer - if len(indexer) > 1: + if len(indexer) > 1 and not multiindex_indexer: l = len(indexer[1]) ser = np.tile(ser, l).reshape(l, -1).T	DataFrame.ix[idx, :] = value sets wrong values when idx is a MultiIndex and DataFrame.columns is also a MultiIndex This code is broken in `0.17.0` but not in `0.15.2`: ``` python import pandas as pd import numpy as np np.random.seed(1) from itertools import product from pandas.util.testing import assert_frame_equal pd.show_versions() idx = pd.MultiIndex.from_tuples( list( product(['A', 'B', 'C'], pd.date_range('2015-01-01', '2015-04-01', freq='MS')) ) ) sub = pd.MultiIndex.from_tuples( [('A', pd.Timestamp('2015-01-01')), ('A', pd.Timestamp('2015-02-01'))] ) # if cols = ['foo', 'bar', 'baz', 'quux'], there is no error. cols = pd.MultiIndex.from_tuples( list( product(['foo', 'bar'], pd.date_range('2015-01-01', '2015-02-01', freq='MS')) ) ) test = pd.DataFrame(np.random.random((12, 4)), index=idx, columns=cols) vals = pd.DataFrame(np.random.random((2, 4)), index=sub, columns=cols) test.ix[sub, :] = vals print test.ix[sub, :] print vals assert_frame_equal(test.ix[sub, :], vals) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.287775 0.130029 0.019367 0.678836 foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.211628 0.265547 0.491573 0.053363 Traceback (most recent call last): File "c:\dev\code\sandbox\multiindex.py", line 41, in <module> assert_frame_equal(test.ix[sub, :], vals) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 1028, in assert_frame_equal obj='DataFrame.iloc[:, {0}]'.format(i)) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 925, in assert_series_equal check_less_precise, obj='{0}'.format(obj)) File "pandas\src\testing.pyx", line 58, in pandas._testing.assert_almost_equal (pandas\src\testing.c:3809) File "pandas\src\testing.pyx", line 147, in pandas._testing.assert_almost_equal (pandas\src\testing.c:2685) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 798, in raise_assert_detail raise AssertionError(msg) AssertionError: DataFrame.iloc[:, 0] are different DataFrame.iloc[:, 0] values are different (50.0 %) [left]: [0.287775338586, 0.287775338586] [right]: [0.287775338586, 0.211628116] ``` ### 0.15.2 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: en_GB pandas: 0.15.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: 0.15.1 statsmodels: None IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.3.0 dateutil: 2.4.1 pytz: 2015.4 bottleneck: 1.0.0 tables: 3.2.0 numexpr: 2.4.3 matplotlib: 1.4.3 openpyxl: 1.8.5 xlrd: 0.9.4 xlwt: 0.7.5 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None rpy2: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.211628 0.265547 0.491573 0.053363 foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.211628 0.265547 0.491573 0.053363 ```	Indexing with a specific set of columns also gives the error: Code sample: ``` python import pandas as pd import numpy as np np.random.seed(1) from itertools import product from pandas.util.testing import assert_frame_equal pd.show_versions() idx = pd.MultiIndex.from_tuples( list( product(['A', 'B', 'C'], pd.date_range('2015-01-01', '2015-04-01', freq='MS')) ) ) cols = pd.MultiIndex.from_tuples( list( product(['foo', 'bar'], pd.date_range('2016-01-01', '2016-02-01', freq='MS')) ) ) # if cols = ['foo', 'bar', 'baz', 'quux'], there is no error. test = pd.DataFrame(np.random.random((12, 4)), index=idx, columns=cols) subidx = pd.MultiIndex.from_tuples( [('A', pd.Timestamp('2015-01-01')), ('A', pd.Timestamp('2015-02-01'))] ) subcols = pd.MultiIndex.from_tuples( [('foo', pd.Timestamp('2016-01-01')), ('foo', pd.Timestamp('2016-02-01'))] ) vals = pd.DataFrame(np.random.random((2, 2)), index=subidx, columns=subcols) test.ix[subidx, subcols] = vals print test.ix[subidx, subcols] print vals assert_frame_equal(test.ix[subidx, subcols], vals) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None foo 2016-01-01 2016-02-01 A 2015-01-01 0.287775 0.130029 2015-02-01 0.287775 0.130029 foo 2016-01-01 2016-02-01 A 2015-01-01 0.287775 0.130029 2015-02-01 0.019367 0.678836 Traceback (most recent call last): File "c:\dev\code\sandbox\multiindex.py", line 48, in <module> assert_frame_equal(test.ix[subidx, subcols], vals) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 1028, in assert_frame_equal obj='DataFrame.iloc[:, {0}]'.format(i)) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 925, in assert_series_equal check_less_precise, obj='{0}'.format(obj)) File "pandas\src\testing.pyx", line 58, in pandas._testing.assert_almost_equal (pandas\src\testing.c:3809) File "pandas\src\testing.pyx", line 147, in pandas._testing.assert_almost_equal (pandas\src\testing.c:2685) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 798, in raise_assert_detail raise AssertionError(msg) AssertionError: DataFrame.iloc[:, 0] are different DataFrame.iloc[:, 0] values are different (50.0 %) [left]: [0.287775338586, 0.287775338586] [right]: [0.287775338586, 0.0193669578703] ``` (Deleted- misread something, my previous suggestion was not really a fix) hmm, surprised that broke. there is not must testing on that sub-section actually The issue is here: https://github.com/pydata/pandas/blob/master/pandas/core/indexing.py#L450 `self._align_series` is called on a sub-section of the frame, but in the aligner, it looks at it and says oh you are a frame so gives back the wrong result. So could prob pass in an additional parameter which would determine this. Since I've already got two test cases, I'd be happy to have a go if I can be pointed in the right direction. I'll start by looking at the history of `indexing.py` and following any referenced issues / PRs the pointer above is to the relevant issues. the way to do this is to setup the test cases and the expected results (in test_indexing); they should fail before a fix, then you can step thru to see where to put a fix and go from there	2015-10-21T08:40:16Z	[]	[]	Traceback (most recent call last): File "c:\dev\code\sandbox\multiindex.py", line 41, in <module> assert_frame_equal(test.ix[sub, :], vals) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 1028, in assert_frame_equal obj='DataFrame.iloc[:, {0}]'.format(i)) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 925, in assert_series_equal check_less_precise, obj='{0}'.format(obj)) File "pandas\src\testing.pyx", line 58, in pandas._testing.assert_almost_equal (pandas\src\testing.c:3809) File "pandas\src\testing.pyx", line 147, in pandas._testing.assert_almost_equal (pandas\src\testing.c:2685) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 798, in raise_assert_detail raise AssertionError(msg) AssertionError: DataFrame.iloc[:, 0] are different	10,878
pandas-dev/pandas	pandas-dev__pandas-11427	faa6cc744ba6086ddcef66c462823a169e1a733c	diff --git a/doc/source/timeseries.rst b/doc/source/timeseries.rst --- a/doc/source/timeseries.rst +++ b/doc/source/timeseries.rst @@ -954,6 +954,52 @@ These can be used as arguments to ``date_range``, ``bdate_range``, constructors for ``DatetimeIndex``, as well as various other timeseries-related functions in pandas. +Anchored Offset Semantics +~~~~~~~~~~~~~~~~~~~~~~~~~ + +For those offsets that are anchored to the start or end of specific +frequency (``MonthEnd``, ``MonthBegin``, ``WeekEnd``, etc) the following +rules apply to rolling forward and backwards. + +When ``n`` is not 0, if the given date is not on an anchor point, it snapped to the next(previous) +anchor point, and moved ``\|n\|-1`` additional steps forwards or backwards. + +.. ipython:: python + + pd.Timestamp('2014-01-02') + MonthBegin(n=1) + pd.Timestamp('2014-01-02') + MonthEnd(n=1) + + pd.Timestamp('2014-01-02') - MonthBegin(n=1) + pd.Timestamp('2014-01-02') - MonthEnd(n=1) + + pd.Timestamp('2014-01-02') + MonthBegin(n=4) + pd.Timestamp('2014-01-02') - MonthBegin(n=4) + +If the given date is on an anchor point, it is moved ``\|n\|`` points forwards +or backwards. + +.. ipython:: python + + pd.Timestamp('2014-01-01') + MonthBegin(n=1) + pd.Timestamp('2014-01-31') + MonthEnd(n=1) + + pd.Timestamp('2014-01-01') - MonthBegin(n=1) + pd.Timestamp('2014-01-31') - MonthEnd(n=1) + + pd.Timestamp('2014-01-01') + MonthBegin(n=4) + pd.Timestamp('2014-01-31') - MonthBegin(n=4) + +For the case when ``n=0``, the date is not moved if on an anchor point, otherwise +it is rolled forward to the next anchor point. + +.. ipython:: python + + pd.Timestamp('2014-01-02') + MonthBegin(n=0) + pd.Timestamp('2014-01-02') + MonthEnd(n=0) + + pd.Timestamp('2014-01-01') + MonthBegin(n=0) + pd.Timestamp('2014-01-31') + MonthEnd(n=0) + .. _timeseries.legacyaliases: Legacy Aliases diff --git a/doc/source/whatsnew/v0.18.0.txt b/doc/source/whatsnew/v0.18.0.txt --- a/doc/source/whatsnew/v0.18.0.txt +++ b/doc/source/whatsnew/v0.18.0.txt @@ -190,7 +190,7 @@ Bug Fixes - + - Bug in vectorized ``DateOffset`` when ``n`` parameter is ``0`` (:issue:`11370`) diff --git a/pandas/tseries/offsets.py b/pandas/tseries/offsets.py --- a/pandas/tseries/offsets.py +++ b/pandas/tseries/offsets.py @@ -444,7 +444,7 @@ def _beg_apply_index(self, i, freq): from pandas.tseries.frequencies import get_freq_code base, mult = get_freq_code(freq) base_period = i.to_period(base) - if self.n < 0: + if self.n <= 0: # when subtracting, dates on start roll to prior roll = np.where(base_period.to_timestamp() == i - off, self.n, self.n + 1) @@ -464,7 +464,7 @@ def _end_apply_index(self, i, freq): base, mult = get_freq_code(freq) base_period = i.to_period(base) if self.n > 0: - # when adding, dtates on end roll to next + # when adding, dates on end roll to next roll = np.where(base_period.to_timestamp(how='end') == i - off, self.n, self.n - 1) else: @@ -1081,8 +1081,7 @@ def apply(self, other): @apply_index_wraps def apply_index(self, i): - months = self.n - 1 if self.n >= 0 else self.n - shifted = tslib.shift_months(i.asi8, months, 'end') + shifted = tslib.shift_months(i.asi8, self.n, 'end') return i._shallow_copy(shifted) def onOffset(self, dt): @@ -1108,8 +1107,7 @@ def apply(self, other): @apply_index_wraps def apply_index(self, i): - months = self.n + 1 if self.n < 0 else self.n - shifted = tslib.shift_months(i.asi8, months, 'start') + shifted = tslib.shift_months(i.asi8, self.n, 'start') return i._shallow_copy(shifted) def onOffset(self, dt): @@ -1777,6 +1775,7 @@ def apply(self, other): @apply_index_wraps def apply_index(self, i): freq_month = 12 if self.startingMonth == 1 else self.startingMonth - 1 + # freq_month = self.startingMonth freqstr = 'Q-%s' % (_int_to_month[freq_month],) return self._beg_apply_index(i, freqstr) diff --git a/pandas/tslib.pyx b/pandas/tslib.pyx --- a/pandas/tslib.pyx +++ b/pandas/tslib.pyx @@ -4458,7 +4458,8 @@ def shift_months(int64_t[:] dtindex, int months, object day=None): Py_ssize_t i pandas_datetimestruct dts int count = len(dtindex) - int days_in_current_month + int months_to_roll + bint roll_check int64_t[:] out = np.empty(count, dtype='int64') if day is None: @@ -4472,36 +4473,44 @@ def shift_months(int64_t[:] dtindex, int months, object day=None): dts.day = min(dts.day, days_in_month(dts)) out[i] = pandas_datetimestruct_to_datetime(PANDAS_FR_ns, &dts) elif day == 'start': + roll_check = False + if months <= 0: + months += 1 + roll_check = True with nogil: for i in range(count): if dtindex[i] == NPY_NAT: out[i] = NPY_NAT; continue pandas_datetime_to_datetimestruct(dtindex[i], PANDAS_FR_ns, &dts) - dts.year = _year_add_months(dts, months) - dts.month = _month_add_months(dts, months) + months_to_roll = months + + # offset semantics - if on the anchor point and going backwards + # shift to next + if roll_check and dts.day == 1: + months_to_roll -= 1 + + dts.year = _year_add_months(dts, months_to_roll) + dts.month = _month_add_months(dts, months_to_roll) + dts.day = 1 - # offset semantics - when subtracting if at the start anchor - # point, shift back by one more month - if months <= 0 and dts.day == 1: - dts.year = _year_add_months(dts, -1) - dts.month = _month_add_months(dts, -1) - else: - dts.day = 1 out[i] = pandas_datetimestruct_to_datetime(PANDAS_FR_ns, &dts) elif day == 'end': + roll_check = False + if months > 0: + months -= 1 + roll_check = True with nogil: for i in range(count): if dtindex[i] == NPY_NAT: out[i] = NPY_NAT; continue pandas_datetime_to_datetimestruct(dtindex[i], PANDAS_FR_ns, &dts) - days_in_current_month = days_in_month(dts) - - dts.year = _year_add_months(dts, months) - dts.month = _month_add_months(dts, months) + months_to_roll = months # similar semantics - when adding shift forward by one # month if already at an end of month - if months >= 0 and dts.day == days_in_current_month: - dts.year = _year_add_months(dts, 1) - dts.month = _month_add_months(dts, 1) + if roll_check and dts.day == days_in_month(dts): + months_to_roll += 1 + + dts.year = _year_add_months(dts, months_to_roll) + dts.month = _month_add_months(dts, months_to_roll) dts.day = days_in_month(dts) out[i] = pandas_datetimestruct_to_datetime(PANDAS_FR_ns, &dts)	Vectorised addition of MonthOffset(n=0) returns different values to item-by-item addition This code returns different values in `0.17.0` and `0.15.2` ``` python import pandas as pd from pandas.util.testing import assert_index_equal pd.show_versions() offsets = [ pd.offsets.Day, pd.offsets.MonthBegin, pd.offsets.QuarterBegin, pd.offsets.YearBegin, ] dates = pd.date_range('2011-01-01', '2011-01-05', freq='D') for offset in offsets: # adding each item individually or vectorised should give same answer expected_vec = dates + offset(n=0) expected = pd.DatetimeIndex([d + offset(n=0) for d in dates]) msg = "offset: {}, vectorised: {}, individual: {}".format( offset, expected_vec, expected ) try: if pd.__version__ == '0.17.0': assert_index_equal(expected_vec, expected, check_names=False) else: assert_index_equal(expected_vec, expected) except AssertionError as er: raise Exception(msg + str(er)) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None Traceback (most recent call last): File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 24, in <module> raise Exception(msg + str(er)) Exception: offset: <class 'pandas.tseries.offsets.MonthBegin'>, vectorised: DatetimeIndex(['2010-12-01', '2011-01-01', '2011-01-01', '2011-01-01', '2011-01-01'], dtype='datetime64[ns]', freq=None), individual: DatetimeIndex(['2011-01-01', '2011-02-01', '2011-02-01', '2011-02-01', '2011-02-01'], dtype='datetime64[ns]', freq=None)Index are different Index values are different (100.0 %) [left]: DatetimeIndex(['2010-12-01', '2011-01-01', '2011-01-01', '2011-01-01', '2011-01-01'], dtype='datetime64[ns]', freq=None) [right]: DatetimeIndex(['2011-01-01', '2011-02-01', '2011-02-01', '2011-02-01', '2011-02-01'], dtype='datetime64[ns]', freq=None) ``` ### 0.15.2 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: en_GB pandas: 0.15.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: 0.15.1 statsmodels: None IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.3.0 dateutil: 2.4.1 pytz: 2015.4 bottleneck: 1.0.0 tables: 3.2.0 numexpr: 2.4.3 matplotlib: 1.4.3 openpyxl: 1.8.5 xlrd: 0.9.4 xlwt: 0.7.5 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None rpy2: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None ```	This is from #10744, I didn't have the n=0 semantics right (and apparently didn't test!). It'll be a couple days, but I'll submit a fix. Many thanks for quick response! `MonthEnd` also not working: Test script ``` python import pandas as pd from pandas.util.testing import assert_index_equal pd.show_versions() offsets = [ pd.offsets.MonthEnd, pd.offsets.QuarterEnd, pd.offsets.YearEnd, ] dates = pd.date_range('2011-01-01', '2011-01-05', freq='D') for offset in offsets: # adding each item individually or vectorised should give same answer expected_vec = dates + offset(n=0) expected = pd.DatetimeIndex([d + offset(n=0) for d in dates]) msg = "offset: {}, vectorised: {}, individual: {}".format( offset, expected_vec, expected ) try: if pd.__version__ == '0.17.0': assert_index_equal(expected_vec, expected, check_names=False) else: assert_index_equal(expected_vec, expected) except AssertionError as er: raise Exception(msg + str(er)) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None Traceback (most recent call last): File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 34, in <module> raise Exception(msg + str(er)) Exception: offset: <class 'pandas.tseries.offsets.MonthEnd'>, vectorised: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31'], dtype='datetime64[ns]', freq=None), individual: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31'], dtype='datetime64[ns]', freq=None)Index are different Index values are different (100.0 %) [left]: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31'], dtype='datetime64[ns]', freq=None) [right]: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31'], dtype='datetime64[ns]', freq=None) ``` ### 0.15.2 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: en_GB pandas: 0.15.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: 0.15.1 statsmodels: None IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.3.0 dateutil: 2.4.1 pytz: 2015.4 bottleneck: 1.0.0 tables: 3.2.0 numexpr: 2.4.3 matplotlib: 1.4.3 openpyxl: 1.8.5 xlrd: 0.9.4 xlwt: 0.7.5 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None rpy2: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None ``` Probably also wrong for `YearEnd` and `QuarterEnd` too as the counting logic is shared IIRC. On Mon, Oct 19, 2015 at 10:31 AM, Petra Chong notifications@github.com wrote: > MonthEnd also not working: > > Test script > > import pandas as pdfrom pandas.util.testing import assert_index_equal > > pd.show_versions() > > offsets = [ > pd.offsets.MonthEnd, > pd.offsets.QuarterEnd, pd.offsets.YearEnd, > ] > > dates = pd.date_range('2011-01-01', '2011-01-05', freq='D') > for offset in offsets: > # adding each item individually or vectorised should give same answer > expected_vec = dates + offset(n=0) > expected = pd.DatetimeIndex([d + offset(n=0) for d in dates]) > > ``` > msg = "offset: {}, vectorised: {}, individual: {}".format( > offset, expected_vec, expected > ) > try: > if pd.__version__ == '0.17.0': > assert_index_equal(expected_vec, expected, check_names=False) > else: > assert_index_equal(expected_vec, expected) > except AssertionError as er: > raise Exception(msg + str(er)) > ``` > > 0.17.0 > > INSTALLED VERSIONS------------------ > commit: None > python: 2.7.10.final.0 > python-bits: 64 > OS: Windows > OS-release: 7 > machine: AMD64 > processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel > byteorder: little > LC_ALL: None > LANG: None > > pandas: 0.17.0 > nose: 1.3.7 > pip: 7.1.0 > setuptools: 18.0.1 > Cython: 0.22 > numpy: 1.10.1 > scipy: 0.16.0 > statsmodels: 0.6.1 > IPython: 3.2.1 > sphinx: 1.3.1 > patsy: 0.4.0 > dateutil: 2.4.1 > pytz: 2015.4 > blosc: None > bottleneck: 1.0.0 > tables: 3.2.2 > numexpr: 2.4.4 > matplotlib: 1.4.3 > openpyxl: None > xlrd: 0.9.4 > xlwt: None > xlsxwriter: 0.7.3 > lxml: None > bs4: 4.3.2 > html5lib: 0.999 > httplib2: None > apiclient: None > sqlalchemy: 1.0.7 > pymysql: None > psycopg2: None > Traceback (most recent call last): > File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 34, in <module> > raise Exception(msg + str(er))Exception: offset: <class 'pandas.tseries.offsets.MonthEnd'>, vectorised: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', > '2010-12-31'], > dtype='datetime64[ns]', freq=None), individual: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', > '2011-01-31'], > dtype='datetime64[ns]', freq=None)Index are different > > Index values are different (100.0 %) > [left]: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', > '2010-12-31'], > dtype='datetime64[ns]', freq=None) > [right]: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', > '2011-01-31'], > dtype='datetime64[ns]', freq=None) > > 0.15.2 > > INSTALLED VERSIONS------------------ > commit: None > python: 2.7.10.final.0 > python-bits: 64 > OS: Windows > OS-release: 7 > machine: AMD64 > processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel > byteorder: little > LC_ALL: None > LANG: en_GB > > pandas: 0.15.2 > nose: 1.3.7 > Cython: 0.22 > numpy: 1.9.2 > scipy: 0.15.1 > statsmodels: None > IPython: 3.2.1 > sphinx: 1.3.1 > patsy: 0.3.0 > dateutil: 2.4.1 > pytz: 2015.4 > bottleneck: 1.0.0 > tables: 3.2.0 > numexpr: 2.4.3 > matplotlib: 1.4.3 > openpyxl: 1.8.5 > xlrd: 0.9.4 > xlwt: 0.7.5 > xlsxwriter: 0.7.3 > lxml: 3.4.4 > bs4: 4.3.2 > html5lib: 0.999 > httplib2: None > apiclient: None > rpy2: None > sqlalchemy: 1.0.7 > pymysql: None > psycopg2: None > > — > Reply to this email directly or view it on GitHub > https://github.com/pydata/pandas/issues/11370#issuecomment-149250402.	2015-10-24T21:09:45Z	[]	[]	Traceback (most recent call last): File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 24, in <module> raise Exception(msg + str(er)) Exception: offset: <class 'pandas.tseries.offsets.MonthBegin'>, vectorised: DatetimeIndex(['2010-12-01', '2011-01-01', '2011-01-01', '2011-01-01',	10,881
pandas-dev/pandas	pandas-dev__pandas-11653	2d038327b7ea805a2d9c3db9ca3dd2b459e694bb	diff --git a/doc/source/whatsnew/v0.17.1.txt b/doc/source/whatsnew/v0.17.1.txt --- a/doc/source/whatsnew/v0.17.1.txt +++ b/doc/source/whatsnew/v0.17.1.txt @@ -169,7 +169,7 @@ Bug Fixes - +- Bug in indexing with a ``range``, (:issue:`11652`) - Bug in ``to_sql`` using unicode column names giving UnicodeEncodeError with (:issue:`11431`). diff --git a/pandas/core/index.py b/pandas/core/index.py --- a/pandas/core/index.py +++ b/pandas/core/index.py @@ -1755,7 +1755,8 @@ def get_loc(self, key, method=None, tolerance=None): if tolerance is not None: raise ValueError('tolerance argument only valid if using pad, ' 'backfill or nearest lookups') - return self._engine.get_loc(_values_from_object(key)) + key = _values_from_object(key) + return self._engine.get_loc(key) indexer = self.get_indexer([key], method=method, tolerance=tolerance) diff --git a/pandas/core/indexing.py b/pandas/core/indexing.py --- a/pandas/core/indexing.py +++ b/pandas/core/indexing.py @@ -104,6 +104,8 @@ def _get_setitem_indexer(self, key): if isinstance(key, tuple) and not self.ndim < len(key): return self._convert_tuple(key, is_setter=True) + if isinstance(key, range): + return self._convert_range(key, is_setter=True) try: return self._convert_to_indexer(key, is_setter=True) @@ -156,6 +158,10 @@ def _convert_tuple(self, key, is_setter=False): keyidx.append(idx) return tuple(keyidx) + def _convert_range(self, key, is_setter=False): + """ convert a range argument """ + return list(key) + def _convert_scalar_indexer(self, key, axis): # if we are accessing via lowered dim, use the last dim ax = self.obj._get_axis(min(axis, self.ndim - 1))	BUG: ValueError when indexing using range only when length >= 1,000,000 From SO (http://stackoverflow.com/questions/33814223/strange-error-in-pandas-indexing-with-range-when-length-1-000-000) Pandas raises a ValueError when assigning multiple values to a Series (or DataFrame) using range(x) where x > 1. This error is raised only when its length is one million or larger. ``` python import pandas as pd for x in [5, 999999, 1000000]: s = pd.Series(index=range(x)) print('series length =', len(s)) # assigning value with range(1), always works s.loc[range(1)] = 42 # reading values with range(x>1), always works _ = s.loc[range(2)] # assigning values with range(x>1), fails only when len >= 1 million s.loc[range(2)] = 42 ``` Output: ``` python series length = 5 series length = 999999 series length = 1000000 Traceback (most recent call last): File "<stdin>", line 9, in <module> File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 114, in __setitem__ indexer = self._get_setitem_indexer(key) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 109, in _get_setitem_indexer return self._convert_to_indexer(key, is_setter=True) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 1042, in _convert_to_indexer return labels.get_loc(obj) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/index.py", line 1692, in get_loc return self._engine.get_loc(_values_from_object(key)) File "pandas/index.pyx", line 137, in pandas.index.IndexEngine.get_loc (pandas/index.c:3979) File "pandas/index.pyx", line 145, in pandas.index.IndexEngine.get_loc (pandas/index.c:3680) File "pandas/index.pyx", line 464, in pandas.index._bin_search (pandas/index.c:9124) ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() ``` Tested on pandas 0.17.0 and python 3.4.	odd, only happens on py3. `range(2)` is a list in py2.x. `s.loc[list[range(2)]] = 42` works fine with py3, too. It seems to be failing only with `range` object.	2015-11-19T22:47:48Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 9, in <module> File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 114, in __setitem__ indexer = self._get_setitem_indexer(key) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 109, in _get_setitem_indexer return self._convert_to_indexer(key, is_setter=True) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 1042, in _convert_to_indexer return labels.get_loc(obj) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/index.py", line 1692, in get_loc return self._engine.get_loc(_values_from_object(key)) File "pandas/index.pyx", line 137, in pandas.index.IndexEngine.get_loc (pandas/index.c:3979) File "pandas/index.pyx", line 145, in pandas.index.IndexEngine.get_loc (pandas/index.c:3680) File "pandas/index.pyx", line 464, in pandas.index._bin_search (pandas/index.c:9124) ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()	10,898
pandas-dev/pandas	pandas-dev__pandas-11714	547750aa5ba5b4a1b5d0cde05cc21e588b30cc27	diff --git a/doc/source/whatsnew/v0.18.0.txt b/doc/source/whatsnew/v0.18.0.txt --- a/doc/source/whatsnew/v0.18.0.txt +++ b/doc/source/whatsnew/v0.18.0.txt @@ -32,6 +32,7 @@ Other enhancements ^^^^^^^^^^^^^^^^^^ - Handle truncated floats in SAS xport files (:issue:`11713`) +- ``read_excel`` now supports s3 urls of the format ``s3://bucketname/filename`` (:issue:`11447`) .. _whatsnew_0180.enhancements.rounding: diff --git a/pandas/io/excel.py b/pandas/io/excel.py --- a/pandas/io/excel.py +++ b/pandas/io/excel.py @@ -11,7 +11,7 @@ from pandas.core.frame import DataFrame from pandas.io.parsers import TextParser -from pandas.io.common import _is_url, _urlopen, _validate_header_arg +from pandas.io.common import _is_url, _urlopen, _validate_header_arg, get_filepath_or_buffer, _is_s3_url from pandas.tseries.period import Period from pandas import json from pandas.compat import (map, zip, reduce, range, lrange, u, add_metaclass, @@ -199,7 +199,10 @@ def __init__(self, io, **kwds): raise ValueError("Unknown engine: %s" % engine) if isinstance(io, compat.string_types): - if _is_url(io): + if _is_s3_url(io): + buffer, _, _ = get_filepath_or_buffer(io) + self.book = xlrd.open_workbook(file_contents=buffer.read()) + elif _is_url(io): data = _urlopen(io).read() self.book = xlrd.open_workbook(file_contents=data) else:	ENH read_excel error when accessing AWS S3 URL Summary: read_excel is unable to read a file using the same S3 URL syntax as read_csv. read_excel should support accessing S3 data in the same manner as read_csv read_excel fails with the following error: ``` python >>> import pandas as pd >>> df = pd.read_excel("s3://my-bucket/my_file.xlsx") Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 163, in read_excel io = ExcelFile(io, engine=engine) File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 206, in __init__ self.book = xlrd.open_workbook(io) File "/usr/local/lib/python2.6/site-packages/xlrd/__init__.py", line 394, in open_workbook f = open(filename, "rb") IOError: [Errno 2] No such file or directory: 's3://my-bucket/my_file.xlsx' >>> ``` read_csv on the other hand is able to successfully read a csv file in the same S3 bucket using the same URL syntax: ``` python >>> import pandas as pd >>> df = pd.read_csv("s3://my-bucket/my_file.csv") >>> len(df.index) 1187 >>> ``` For the record, read_csv can also see the xlsx file but returns parse errors when attempting to tokenize the data. ``` python >>> import pandas as pd >>> df = pd.read_csv("s3://my-bucket/my_file.xlsx") Exception pandas.parser.CParserError: CParserError('Error tokenizing data. C error: Expected 9 fields in line 210, saw 10\n',) in 'pandas.parser.TextReader._tokenize_rows' ignored >>> ``` read_excel successfully reads and parses a local copy of the xlsx file ``` python >>> import pandas as pd >>> df = pd.read_excel("my_file.xlsx") >>> len(df.index) 221 >>> ``` Pandas version string and dependencies: ``` python >>> pd.show_versions() INSTALLED VERSIONS ------------------ commit: None python: 2.6.9.final.0 python-bits: 64 OS: Linux OS-release: 3.14.48-33.39.amzn1.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 pandas: 0.17.0 nose: 1.3.4 pip: 6.1.1 setuptools: 12.2 Cython: None numpy: 1.10.1 scipy: 0.16.0 statsmodels: None IPython: None sphinx: None patsy: None dateutil: 2.4.2 pytz: 2015.7 blosc: None bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None >>> ```	This is almost a trivial enhancement, just add `_is_s3_url` [here](https://github.com/pydata/pandas/blob/master/pandas/io/excel.py#L202) post a file link as an example and i'll put it on our test s3 bucket.	2015-11-27T20:09:28Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 163, in read_excel io = ExcelFile(io, engine=engine) File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 206, in __init__ self.book = xlrd.open_workbook(io) File "/usr/local/lib/python2.6/site-packages/xlrd/__init__.py", line 394, in open_workbook f = open(filename, "rb") IOError: [Errno 2] No such file or directory: 's3://my-bucket/my_file.xlsx'	10,901
pandas-dev/pandas	pandas-dev__pandas-12043	1ae6384a0c04be8a1faddaa85751a9cac2f5a42a	diff --git a/pandas/tseries/index.py b/pandas/tseries/index.py --- a/pandas/tseries/index.py +++ b/pandas/tseries/index.py @@ -1804,7 +1804,6 @@ def indexer_between_time(self, start_time, end_time, include_start=True, "%I%M%S%p") include_start : boolean, default True include_end : boolean, default True - tz : string or pytz.timezone or dateutil.tz.tzfile, default None Returns -------	Error in doc of DatetimeIndex.indexer_between_time Hello everyone, This is not a major issue: I was trying to use the `tz` parameter as indicated on the [documentation](http://pandas.pydata.org/pandas-docs/version/0.17.1/generated/pandas.DatetimeIndex.indexer_between_time.html?highlight=indexer_between_time#pandas.DatetimeIndex.indexer_between_time) of `DatetimeIndex.indexer_between_time` and it turns out that `tz` is not implemented. Here is an example: ``` python import pandas as pd pd.DatetimeIndex(['2016-01-01 00:00:00', '2016-01-01 01:00:00', '2016-01-01 02:00:00']).indexer_between_time('01:00', '02:00', tz='Europe/Paris') ``` ``` Traceback (most recent call last): File "~/anaconda/lib/python2.7/site-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-98-258497b42264>", line 1, in <module> pd.DatetimeIndex(['2016-01-01 00:00:00', '2016-01-01 01:00:00', '2016-01-01 02:00:00']).indexer_between_time('01:00', '02:00', tz='Europe/Paris') TypeError: indexer_between_time() got an unexpected keyword argument 'tz' ```	That doc should be fixed (and you see the signature no longer has a tz argument). Having a `tz` makes no sense in this context. @joseRLC want to do a pull-request for that? Can I try this out? This seems like a good issue for a first PR. @RahulHP Certainly, go for it! If you have any questions about the workflow (see http://pandas.pydata.org/pandas-docs/stable/contributing.html), just ask (you can also use the gitter channel for that: https://gitter.im/pydata/pandas)	2016-01-15T08:44:51Z	[]	[]	Traceback (most recent call last): File "~/anaconda/lib/python2.7/site-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-98-258497b42264>", line 1, in <module> pd.DatetimeIndex(['2016-01-01 00:00:00', '2016-01-01 01:00:00', '2016-01-01 02:00:00']).indexer_between_time('01:00', '02:00', tz='Europe/Paris') TypeError: indexer_between_time() got an unexpected keyword argument 'tz'	10,929
pandas-dev/pandas	pandas-dev__pandas-12058	1945eed731a9d8fdb9a21837b326c42f8771def7	diff --git a/doc/source/whatsnew/v0.18.0.txt b/doc/source/whatsnew/v0.18.0.txt --- a/doc/source/whatsnew/v0.18.0.txt +++ b/doc/source/whatsnew/v0.18.0.txt @@ -445,7 +445,7 @@ Bug Fixes - Accept unicode in ``Timedelta`` constructor (:issue:`11995`) - Bug in value label reading for ``StataReader`` when reading incrementally (:issue:`12014`) - Bug in vectorized ``DateOffset`` when ``n`` parameter is ``0`` (:issue:`11370`) - +- Compat for numpy 1.11 w.r.t. ``NaT`` comparison changes (:issue:`12049`) diff --git a/pandas/core/common.py b/pandas/core/common.py --- a/pandas/core/common.py +++ b/pandas/core/common.py @@ -379,12 +379,13 @@ def array_equivalent(left, right, strict_nan=False): """ left, right = np.asarray(left), np.asarray(right) + + # shape compat if left.shape != right.shape: return False # Object arrays can contain None, NaN and NaT. - if (issubclass(left.dtype.type, np.object_) or - issubclass(right.dtype.type, np.object_)): + if is_object_dtype(left) or is_object_dtype(right): if not strict_nan: # pd.isnull considers NaN and None to be equivalent. @@ -405,13 +406,21 @@ def array_equivalent(left, right, strict_nan=False): return True # NaNs can occur in float and complex arrays. - if issubclass(left.dtype.type, (np.floating, np.complexfloating)): + if is_float_dtype(left) or is_complex_dtype(left): return ((left == right) \| (np.isnan(left) & np.isnan(right))).all() # numpy will will not allow this type of datetimelike vs integer comparison elif is_datetimelike_v_numeric(left, right): return False + # M8/m8 + elif needs_i8_conversion(left) and needs_i8_conversion(right): + if not is_dtype_equal(left.dtype, right.dtype): + return False + + left = left.view('i8') + right = right.view('i8') + # NaNs cannot occur otherwise. return np.array_equal(left, right)	BLD: numpy master changes breaking (24 hrs ago) good build: https://travis-ci.org/pydata/pandas/jobs/102356098: 1.11.0.dev0+51d2ecd (1 hr ago) breaking lots of things: https://travis-ci.org/pydata/pandas/jobs/102596904: 1.11.0.dev0+aa6335c @shoyer IIRC a couple of your PR's were merged in the last day. here's an example: ``` ====================================================================== FAIL: test_coercion_with_setitem_and_series (pandas.tests.test_indexing.TestSeriesNoneCoercion) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/test_indexing.py", line 5247, in test_coercion_with_setitem_and_series expected_series.values, strict_nan=True) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 866, in assert_numpy_array_equal raise_assert_detail(obj, msg, left, right) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 825, in raise_assert_detail raise AssertionError(msg) AssertionError: numpy array are different numpy array values are different (33.33333 %) [left]: [NaT, 2000-01-02T00:00:00.000000000+0000, 2000-01-03T00:00:00.000000000+0000] [right]: [NaT, 2000-01-02T00:00:00.000000000+0000, 2000-01-03T00:00:00.000000000+0000] ```	my guess is that we now need to compare `M8[ns]` by `.view('i8')` then comparing as the `NaT` will then compare equal. Yep, if this is datetime64 related it's my fault :). I though we were already using .view('i8') before making comparisons but I guess I was wrong. If so (especially if this breaks user facing stuff) we may need to hold off on these numpy fixes for a little longer (deprecation cycle?). Sigh... On Fri, Jan 15, 2016 at 7:27 AM, Jeff Reback notifications@github.com wrote: > ## my guess is that we now need to compare `M8[ns]` by `.view('i8')` then comparing as the `NaT` will then compare equal. > > Reply to this email directly or view it on GitHub: > https://github.com/pydata/pandas/issues/12049#issuecomment-171991205 we can adapt. do you want me to open an issue on numpy? ``` In [1]: np.__version__ Out[1]: '1.10.2' In [5]: arr = np.array([np.nan]) In [6]: np.array_equal(arr,arr) Out[6]: False In [7]: arr = np.array([np.datetime64('NaT')]) In [8]: np.array_equal(arr,arr) Out[8]: True ``` ok so your change makes sense ``` In [1]: np.__version__ Out[1]: '1.11.0.dev0+aa6335c' In [2]: In [5]: arr = np.array([np.nan]) In [7]: arr = np.array([np.nan]) In [8]: np.array_equal(arr,arr) Out[8]: False In [9]: arr = np.array([np.datetime64('NaT')]) In [10]: np.array_equal(arr,arr) Out[10]: False ``` Actually, looks like we might just be able to drop our `assert_numpy_array_equivalent` entirely in favor of numpy's `assert_array_equal`, which has supported NaN equality since at least August 2011 (numpy 1.7, I think): https://github.com/numpy/numpy/commit/67ece6bdd2b35d011893e78154dbff6ab51c7d35 Unfortunately, this bug exists in our `array_equivalent` utility function, which we use for the `equals` method. This means that with this change datetime64 equality checks involving NaT will be broken. As much as I would love to just roll out the NumPy fix, doing it like this will assuredly result in unhappy users and unnecessary aggravation when sanity checks and test suites fail. To fix this, I propose: - We roll back the NaT comparison fix in NumPy for now, issuing a deprecation warning instead for a numpy release or two. - We add the fix to `array_equivalent` in pandas, doing the appropriate cast to int64 to avoid needing to catch the deprecation warning (which can have performance consequences). yep, prob need a special case for `M8/m8` (alternatively we could have a conditional check on the numpy version). Ok, lmk if you need anything on the numpy side.	2016-01-15T20:59:07Z	[]	[]	Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/test_indexing.py", line 5247, in test_coercion_with_setitem_and_series expected_series.values, strict_nan=True) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 866, in assert_numpy_array_equal raise_assert_detail(obj, msg, left, right) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 825, in raise_assert_detail raise AssertionError(msg) AssertionError: numpy array are different	10,930
pandas-dev/pandas	pandas-dev__pandas-13188	3944a369265f27268d1b3867a161e97f9c63cd62	diff --git a/pandas/tools/plotting.py b/pandas/tools/plotting.py --- a/pandas/tools/plotting.py +++ b/pandas/tools/plotting.py @@ -3353,7 +3353,8 @@ def _subplots(naxes=None, sharex=False, sharey=False, squeeze=True, if sharex or sharey: warnings.warn("When passing multiple axes, sharex and sharey " "are ignored. These settings must be specified " - "when creating axes", UserWarning) + "when creating axes", UserWarning, + stacklevel=4) if len(ax) == naxes: fig = ax[0].get_figure() return fig, ax @@ -3370,7 +3371,8 @@ def _subplots(naxes=None, sharex=False, sharey=False, squeeze=True, return fig, _flatten(ax) else: warnings.warn("To output multiple subplots, the figure containing " - "the passed axes is being cleared", UserWarning) + "the passed axes is being cleared", UserWarning, + stacklevel=4) fig.clear() nrows, ncols = _get_layout(naxes, layout=layout, layout_type=layout_type)	FAIL: test_scatter_matrix_axis (pandas.tests.test_graphics_others.TestDataFramePlots) in 0.18.1 with py27 When run alone, `test_scatter_matrix_axis` passes. When run together with the rest of `TestDataFramePlots`, it fails. I cannot test with py34 because matplotlib is not available on FreeBSD under py3k. #### Code Sample, a copy-pastable example if possible ``` % nosetests-2.7 pandas.tests.test_graphics_others:TestDataFramePlots.test_scatter_matrix_axis . ---------------------------------------------------------------------- Ran 1 test in 8.150s OK % nosetests-2.7 pandas.tests.test_graphics_others:TestDataFramePlots ..S/usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/matplotlib/artist.py:221: MatplotlibDeprecationWarning: This has been deprecated in mpl 1.5, please use the axes property. A removal date has not been set. warnings.warn(_get_axes_msg, mplDeprecation, stacklevel=1) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) ......F. ====================================================================== FAIL: test_scatter_matrix_axis (pandas.tests.test_graphics_others.TestDataFramePlots) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/local/lib/python2.7/site-packages/pandas/tests/test_graphics_others.py", line 431, in test_scatter_matrix_axis frame=df, range_padding=.1) File "/usr/local/lib/python2.7/contextlib.py", line 24, in __exit__ self.gen.next() File "/usr/local/lib/python2.7/site-packages/pandas/util/testing.py", line 2318, in assert_produces_warning % expected_warning.__name__) AssertionError: Did not see expected warning of class 'UserWarning'. ---------------------------------------------------------------------- Ran 11 tests in 74.896s FAILED (SKIP=1, failures=1) ``` #### Expected Output I expect all tests to pass. #### output of `pd.show_versions()` ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.11.final.0 python-bits: 64 OS: FreeBSD OS-release: 10.2-STABLE machine: amd64 processor: amd64 byteorder: little LC_ALL: None LANG: None pandas: 0.18.1 nose: 1.3.7 pip: 8.0.2 setuptools: 20.0 Cython: None numpy: 1.11.0 scipy: 0.16.1 statsmodels: None xarray: None IPython: None sphinx: None patsy: None dateutil: 2.5.0 pytz: 2016.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.5.2 matplotlib: 1.5.0 openpyxl: 2.3.5 xlrd: 0.9.3 xlwt: 0.7.5 xlsxwriter: 0.8.5 lxml: 3.5.0 bs4: 4.4.1 html5lib: 0.9999999 httplib2: None apiclient: None sqlalchemy: 0.7.10 pymysql: None psycopg2: None jinja2: 2.8 boto: 2.39.0 pandas_datareader: None ```	Thanks. Haven't been able to reproduce yet, but I'm going to clean up that testing module anyway to catch all those userwarnings and fix the matplotlib deprecation warning. Hopefully I'll figure out what's going wrong.	2016-05-15T17:11:08Z	[]	[]	Traceback (most recent call last): File "/usr/local/lib/python2.7/site-packages/pandas/tests/test_graphics_others.py", line 431, in test_scatter_matrix_axis frame=df, range_padding=.1) File "/usr/local/lib/python2.7/contextlib.py", line 24, in __exit__ self.gen.next() File "/usr/local/lib/python2.7/site-packages/pandas/util/testing.py", line 2318, in assert_produces_warning % expected_warning.__name__) AssertionError: Did not see expected warning of class 'UserWarning'.	10,937
pandas-dev/pandas	pandas-dev__pandas-13641	20de2661c8eff66e465248cbe28062eae0e0e3bb		Test failure with matplotlib 1.5.2rc2 on Debian When running the test suite after the package was built, we get the following failure: ``` ERROR: test_plot (pandas.tests.test_graphics.TestDataFramePlots) ---------------------------------------------------------------------- Traceback (most recent call last): File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tests/test_graphics.py", line 1322, in test_plot df.plot.line(blarg=True) File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 3758, in line return self(kind='line', x=x, y=y, *kwds) [...] File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 1340, in _plot return ax.plot(args, *kwds) File "/usr/lib/python2.7/dist-packages/matplotlib/__init__.py", line 1821, in inner return func(ax, args, *kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_axes.py", line 1432, in plot for line in self._get_lines(args, kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 381, in _grab_next_args for seg in self._plot_args(remaining, kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 369, in _plot_args seg = func(x[:, j % ncx], y[:, j % ncy], kw, kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 276, in _makeline seg = mlines.Line2D(x, y, kw) File "/usr/lib/python2.7/dist-packages/matplotlib/lines.py", line 380, in __init__ self.update(kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/artist.py", line 859, in update raise AttributeError('Unknown property %s' % k) AttributeError: Unknown property blarg ``` Full log [here](https://tests.reproducible-builds.org/debian/rbuild/testing/amd64/pandas_0.18.0+git114-g6c692ae-1.rbuild.log) (from the [Debian bug report](https://bugs.debian.org/827938)). `pandas.show_versions()` gives: ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.12.final.0 python-bits: 64 OS: Linux OS-release: 4.6.0-1-amd64 machine: x86_64 processor: byteorder: little LC_ALL: C LANG: C pandas: 0.18.0+git114-g6c692ae nose: 1.3.7 pip: None setuptools: 20.10.1 Cython: None numpy: 1.11.1rc1 scipy: 0.17.1 statsmodels: 0.6.1 xarray: None IPython: None sphinx: 1.4.4 patsy: 0.4.1 dateutil: 2.4.2 pytz: 2015.7 blosc: None bottleneck: None tables: 3.2.2 numexpr: 2.6.0 matplotlib: 1.5.2rc2 openpyxl: 2.3.0 xlrd: 1.0.0 xlwt: 0.7.5 xlsxwriter: None lxml: 3.6.0 bs4: 4.4.1 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None jinja2: 2.8 boto: None pandas_datareader: None ```	Thanks for the report. Looks to be caused by mpl which raises `AttributeError` rather than `TypeError`. Should change the test on pandas side. - https://github.com/pydata/pandas/blob/master/pandas/tests/test_graphics.py#L1317 Appreciated to submit a PR. [line 1333](https://github.com/pydata/pandas/blob/master/pandas/tests/test_graphics.py#L1333) Shall this be compatible with matplotlib < 1.5.2rc? Then this is not that simple, since tm.assertRaises() only accepts a singe exception, and one then would need to create another mpl_ge_1_5_2 comparison specifically for this test. Is this really worth it? What is the rationale of the test? cc @tacaswell Just to be sure, the change from TypeError to AttributeError, was this on purpose? (it's raised in `set_lineprops`) I do not think that was intentional and I am not sure off the top of my head why this changed. The `Artist.update` method has raised `AttributeError` since 2004. So, I broke this: https://github.com/matplotlib/matplotlib/pull/6175 It looks like we used to have `Artist.set` which raised `TypeError` and `Artist.update` which raised `AttributeError` (because history). These two code-paths got merged in https://github.com/matplotlib/matplotlib/pull/5599 (also my fault) and we missed the API change in the the exceptions. Unfortunately I _just_ tagged 1.5.2 last weekend, but have not posted it to pypi or publicized it yet :disappointed:. I am inclined to just document this as an API change, but if you want to lobby for a 1.5.3 fixing this, this is the time to do it! I don't think it is that an important change, so leaving (and documenting) it is fine for me (for pandas it is only a test that is broken, and I think you will have a better idea of how big this change is for matploltib users). In any case, I think the AttributeError is more logical. @sinhrks @jorisvandenbossche For me the question here is still, what that test is for? can u add `_mpl_ge_1_5_2()` function like below, then: ``` if _mpl_ge_1_5_2(): with tm.assertRaises(AttributeError): ... else: with ... ``` - https://github.com/pydata/pandas/blob/master/pandas/tools/plotting.py#L129 > For me the question here is still, what that test is for? Well, it tests that if you pass a wrong keyword, that this raises an error, and is not swallowed somewhere in the implementation. So it's not a huge important test, but still useful (we actually have to many functions in pandas that silently swallow invalid arguments)	2016-07-13T15:01:57Z	[]	[]	Traceback (most recent call last): File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tests/test_graphics.py", line 1322, in test_plot df.plot.line(blarg=True) File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 3758, in line return self(kind='line', x=x, y=y, *kwds) [...] File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 1340, in _plot return ax.plot(args, *kwds) File "/usr/lib/python2.7/dist-packages/matplotlib/__init__.py", line 1821, in inner return func(ax, args, *kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_axes.py", line 1432, in plot for line in self._get_lines(args, kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 381, in _grab_next_args for seg in self._plot_args(remaining, kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 369, in _plot_args seg = func(x[:, j % ncx], y[:, j % ncy], kw, kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 276, in _makeline seg = mlines.Line2D(x, y, kw) File "/usr/lib/python2.7/dist-packages/matplotlib/lines.py", line 380, in __init__ self.update(kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/artist.py", line 859, in update raise AttributeError('Unknown property %s' % k) AttributeError: Unknown property blarg	10,954
pandas-dev/pandas	pandas-dev__pandas-14126	47a8e713430707afcfe76e7ca995902628d4bccf	diff --git a/pandas/util/print_versions.py b/pandas/util/print_versions.py --- a/pandas/util/print_versions.py +++ b/pandas/util/print_versions.py @@ -101,7 +101,10 @@ def show_versions(as_json=False): deps_blob = list() for (modname, ver_f) in deps: try: - mod = importlib.import_module(modname) + if modname in sys.modules: + mod = sys.modules[modname] + else: + mod = importlib.import_module(modname) ver = ver_f(mod) deps_blob.append((modname, ver)) except:	DataFrame.__repr__ raises TypeError after pd.show_versions() was run Maybe one of the imports in `show_versions` has unwanted side effects? ``` python >>> import pandas as pd >>> pd.DataFrame({'spam': range(10)}) spam 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 >>> pd.show_versions() INSTALLED VERSIONS ------------------ commit: None python: 2.7.12.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 42 Stepping 7, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.18.1 nose: 1.3.7 pip: 8.1.2 setuptools: 24.0.3 Cython: 0.24.1 numpy: 1.11.1 scipy: 0.18.0rc2 statsmodels: None xarray: None IPython: 5.0.0 sphinx: 1.4.5 patsy: 0.4.1 dateutil: 2.5.3 pytz: 2016.6.1 blosc: None bottleneck: 1.1.0 tables: None numexpr: 2.6.0 matplotlib: 1.5.1 openpyxl: 2.3.5 xlrd: 1.0.0 xlwt: None xlsxwriter: None lxml: 3.6.0 bs4: None html5lib: 0.999999999 httplib2: 0.9.2 apiclient: None sqlalchemy: 1.0.14 pymysql: None psycopg2: 2.6.2 (dt dec pq3 ext lo64) jinja2: 2.8 boto: None pandas_datareader: None >>> pd.DataFrame({'spam': range(10)}) Traceback (most recent call last): File "<pyshell#3>", line 1, in <module> pd.DataFrame({'spam': range(10)}) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 67, in __repr__ return str(self) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 47, in __str__ return self.__bytes__() File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 59, in __bytes__ return self.__unicode__().encode(encoding, 'replace') File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 535, in __unicode__ line_width=width, show_dimensions=show_dimensions) File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 1488, in to_string formatter.to_string() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 549, in to_string strcols = self._to_str_columns() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 467, in _to_str_columns str_index = self._get_formatted_index(frame) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 746, in _get_formatted_index fmt_index = [index.format(name=show_index_names, formatter=fmt)] File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1462, in format return self._format_with_header(header, kwargs) File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1486, in _format_with_header result = _trim_front(format_array(values, None, justify='left')) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2007, in format_array return fmt_obj.get_result() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2027, in get_result return _make_fixed_width(fmt_values, self.justify) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2394, in _make_fixed_width max_len = np.max([adj.len(x) for x in strings]) File "C:\Program Files\Python27\lib\site-packages\numpy\core\fromnumeric.py", line 2293, in amax out=out, kwargs) File "C:\Program Files\Python27\lib\site-packages\numpy\core\_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) TypeError: an integer is required ```	I cannot reproduce this using Windows and python 2.7. Could you try to debug this? To see where the error is coming from (seems there is something wrong with the `np.max([adj.len(x) for x in strings])`). Or eg create an isolated environment with only required dependencies to see of the problem occurs there as well (using conda or virtualenv). Tracked it to [this import](https://github.com/pydata/pandas/commit/b4e2d34edcbc404f6c90f76b67bcc5fe26f0945f#diff-24212510f4a09e0461c2b6754d34626dL103) of `numpy`, which according to [the docs](https://docs.python.org/2/library/imp.html#imp.load_module) does a `reload()`. Indeed `numpy` (at least on my machines) seems to dislike being reloaded: ``` python >>> import numpy as np >>> np.max([42]) 42 >>> reload(np) <module 'numpy' from 'C:\Program Files\Python27\lib\site-packages\numpy\__init__.pyc'> >>> np.max([42]) Traceback (most recent call last): File "<pyshell#3>", line 1, in <module> np.max([42]) File "C:\Program Files\Python27\lib\site-packages\numpy\core\fromnumeric.py", line 2293, in amax out=out, **kwargs) File "C:\Program Files\Python27\lib\site-packages\numpy\core\_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) TypeError: an integer is required ``` The line was changed in b4e2d34edcbc404f6c90f76b67bcc5fe26f0945f, so I guess that should fix this for the next version (though I am still curious if others have this `numpy` issue with `reload`). I can't reproduce it using NumPy 1.10.4 and 1.11.1. Can you report it to NumPy? Seems to be a problem with the [binaries here](http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy) (does not occur with the PyPI wheels): I'll report to the packager, closing. The numpy issue: numpy/numpy#7844 @xflr6 Thanks for tracking it down! I think there is a real pandas bug here. The bug is that `show_versions` calls [`importlib.import_module`](https://github.com/pydata/pandas/blob/b4e2d34edcbc404f6c90f76b67bcc5fe26f0945f/pandas/util/print_versions.py#L102), and apparently -- this is not documented anywhere, and may vary between py2 and py3 -- `import_module` may reload modules. `show_versions` should not be reloading all these modules. I'd suggest replacing that line with something like ``` python if modname in sys.modules: mod = sys.modules[modname] else: mod = importlib.import_module(modname) ``` @njsmith but seems numpy is not robust to being reloaded. @jreback I suspect a lot of modules are not robust against reloading. For instance if you define a class in `foo.py`, use it in `bar.py`, instanciate `a = foo.MyClass()` in another module, then use `isinstance(a, foo.MyClass)` that statement will fail if `foo` is reloaded. I suspect what is wanted in many cases is a simple `import`, but I haven't checked. Numpy also uses `load_module` in a few places that should probably be audited. @njsmith The reload property is documented in the `imp` module documentation. It always happens for existing modules. Note that nonexisting modules get created... ok this should be easy to fix then AFAICT, `load_module` is useful when you need to use a module that is not installed and not located in the current directory. For instance, during the numpy install process. If numpy is installed you should be able to simply import it. Maybe `__import__`? Or `importlib.import_module` Pandas actually uses `importlib.import_module`, which isn't documented to reload, but I guess it must eventually call `load_module` because otherwise we wouldn't have this problem. (I haven't tried tracing the details, and `importlib` has completely different implementations on different versions of python, so that's something to watch out for if anyone wants to figure out exactly what's happening).	2016-08-31T10:35:26Z	[]	[]	Traceback (most recent call last): File "<pyshell#3>", line 1, in <module> pd.DataFrame({'spam': range(10)}) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 67, in __repr__ return str(self) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 47, in __str__ return self.__bytes__() File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 59, in __bytes__ return self.__unicode__().encode(encoding, 'replace') File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 535, in __unicode__ line_width=width, show_dimensions=show_dimensions) File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 1488, in to_string formatter.to_string() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 549, in to_string strcols = self._to_str_columns() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 467, in _to_str_columns str_index = self._get_formatted_index(frame) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 746, in _get_formatted_index fmt_index = [index.format(name=show_index_names, formatter=fmt)] File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1462, in format return self._format_with_header(header, kwargs) File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1486, in _format_with_header result = _trim_front(format_array(values, None, justify='left')) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2007, in format_array return fmt_obj.get_result() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2027, in get_result return _make_fixed_width(fmt_values, self.justify) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2394, in _make_fixed_width max_len = np.max([adj.len(x) for x in strings]) File "C:\Program Files\Python27\lib\site-packages\numpy\core\fromnumeric.py", line 2293, in amax out=out, kwargs) File "C:\Program Files\Python27\lib\site-packages\numpy\core\_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) TypeError: an integer is required	11,001
pandas-dev/pandas	pandas-dev__pandas-14208	5e2f9da6e8e713bd89cfe8760e63583ea7d29879	diff --git a/doc/source/basics.rst b/doc/source/basics.rst --- a/doc/source/basics.rst +++ b/doc/source/basics.rst @@ -188,6 +188,32 @@ And similarly for ``axis="items"`` and ``axis="minor"``. match the broadcasting behavior of Panel. Though it would require a transition period so users can change their code... +Series and Index also support the :func:`divmod` builtin. This function takes +the floor division and modulo operation at the same time returning a two-tuple +of the same type as the left hand side. For example: + +.. ipython:: python + + s = pd.Series(np.arange(10)) + s + div, rem = divmod(s, 3) + div + rem + + idx = pd.Index(np.arange(10)) + idx + div, rem = divmod(idx, 3) + div + rem + +We can also do elementwise :func:`divmod`: + +.. ipython:: python + + div, rem = divmod(s, [2, 2, 3, 3, 4, 4, 5, 5, 6, 6]) + div + rem + Missing data / operations with fill values ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ diff --git a/doc/source/whatsnew/v0.19.0.txt b/doc/source/whatsnew/v0.19.0.txt --- a/doc/source/whatsnew/v0.19.0.txt +++ b/doc/source/whatsnew/v0.19.0.txt @@ -1328,6 +1328,9 @@ Other API Changes - ``pd.read_csv()`` in the C engine will now issue a ``ParserWarning`` or raise a ``ValueError`` when ``sep`` encoded is more than one character long (:issue:`14065`) - ``DataFrame.values`` will now return ``float64`` with a ``DataFrame`` of mixed ``int64`` and ``uint64`` dtypes, conforming to ``np.find_common_type`` (:issue:`10364`, :issue:`13917`) - ``pd.read_stata()`` can now handle some format 111 files, which are produced by SAS when generating Stata dta files (:issue:`11526`) +- ``Series`` and ``Index`` now support ``divmod`` which will return a tuple of + series or indices. This behaves like a standard binary operator with regards + to broadcasting rules (:issue:`14208`). .. _whatsnew_0190.deprecations: diff --git a/pandas/core/ops.py b/pandas/core/ops.py --- a/pandas/core/ops.py +++ b/pandas/core/ops.py @@ -39,7 +39,8 @@ def _create_methods(arith_method, comp_method, bool_method, - use_numexpr, special=False, default_axis='columns'): + use_numexpr, special=False, default_axis='columns', + have_divmod=False): # creates actual methods based upon arithmetic, comp and bool method # constructors. @@ -127,6 +128,15 @@ def names(x): names('ror_'), op('\|')), rxor=bool_method(lambda x, y: operator.xor(y, x), names('rxor'), op('^')))) + if have_divmod: + # divmod doesn't have an op that is supported by numexpr + new_methods['divmod'] = arith_method( + divmod, + names('divmod'), + None, + default_axis=default_axis, + construct_result=_construct_divmod_result, + ) new_methods = dict((names(k), v) for k, v in new_methods.items()) return new_methods @@ -156,7 +166,7 @@ def add_methods(cls, new_methods, force, select, exclude): def add_special_arithmetic_methods(cls, arith_method=None, comp_method=None, bool_method=None, use_numexpr=True, force=False, select=None, - exclude=None): + exclude=None, have_divmod=False): """ Adds the full suite of special arithmetic methods (``__add__``, ``__sub__``, etc.) to the class. @@ -177,6 +187,9 @@ def add_special_arithmetic_methods(cls, arith_method=None, if passed, only sets functions with names in select exclude : iterable of strings (optional) if passed, will not set functions with names in exclude + have_divmod : bool, (optional) + should a divmod method be added? this method is special because it + returns a tuple of cls instead of a single element of type cls """ # in frame, special methods have default_axis = None, comp methods use @@ -184,7 +197,7 @@ def add_special_arithmetic_methods(cls, arith_method=None, new_methods = _create_methods(arith_method, comp_method, bool_method, use_numexpr, default_axis=None, - special=True) + special=True, have_divmod=have_divmod) # inplace operators (I feel like these should get passed an `inplace=True` # or just be removed @@ -618,8 +631,22 @@ def _align_method_SERIES(left, right, align_asobject=False): return left, right +def _construct_result(left, result, index, name, dtype): + return left._constructor(result, index=index, name=name, dtype=dtype) + + +def _construct_divmod_result(left, result, index, name, dtype): + """divmod returns a tuple of like indexed series instead of a single series. + """ + constructor = left._constructor + return ( + constructor(result[0], index=index, name=name, dtype=dtype), + constructor(result[1], index=index, name=name, dtype=dtype), + ) + + def _arith_method_SERIES(op, name, str_rep, fill_zeros=None, default_axis=None, - eval_kwargs): + construct_result=_construct_result, eval_kwargs): """ Wrapper function for Series arithmetic operations, to avoid code duplication. @@ -692,8 +719,14 @@ def wrapper(left, right, name=name, na_op=na_op): lvalues = lvalues.values result = wrap_results(safe_na_op(lvalues, rvalues)) - return left._constructor(result, index=left.index, - name=name, dtype=dtype) + return construct_result( + left, + result, + index=left.index, + name=name, + dtype=dtype, + ) + return wrapper @@ -933,6 +966,10 @@ def wrapper(self, other): 'desc': 'Integer division', 'reversed': False, 'reverse': 'rfloordiv'}, + 'divmod': {'op': 'divmod', + 'desc': 'Integer division and modulo', + 'reversed': False, + 'reverse': None}, 'eq': {'op': '==', 'desc': 'Equal to', @@ -1033,7 +1070,8 @@ def flex_wrapper(self, other, level=None, fill_value=None, axis=0): series_special_funcs = dict(arith_method=_arith_method_SERIES, comp_method=_comp_method_SERIES, - bool_method=_bool_method_SERIES) + bool_method=_bool_method_SERIES, + have_divmod=True) _arith_doc_FRAME = """ Binary operator %s with support to substitute a fill_value for missing data in diff --git a/pandas/indexes/base.py b/pandas/indexes/base.py --- a/pandas/indexes/base.py +++ b/pandas/indexes/base.py @@ -3426,7 +3426,7 @@ def _validate_for_numeric_binop(self, other, op, opstr): def _add_numeric_methods_binary(cls): """ add in numeric methods """ - def _make_evaluate_binop(op, opstr, reversed=False): + def _make_evaluate_binop(op, opstr, reversed=False, constructor=Index): def _evaluate_numeric_binop(self, other): from pandas.tseries.offsets import DateOffset @@ -3448,7 +3448,7 @@ def _evaluate_numeric_binop(self, other): attrs = self._maybe_update_attributes(attrs) with np.errstate(all='ignore'): result = op(values, other) - return Index(result, attrs) + return constructor(result, attrs) return _evaluate_numeric_binop @@ -3478,6 +3478,15 @@ def _evaluate_numeric_binop(self, other): cls.__rdiv__ = _make_evaluate_binop( operator.div, '__div__', reversed=True) + cls.__divmod__ = _make_evaluate_binop( + divmod, + '__divmod__', + constructor=lambda result, attrs: ( + Index(result[0], attrs), + Index(result[1], **attrs), + ), + ) + @classmethod def _add_numeric_methods_unary(cls): """ add in numeric unary methods """	Regression: divmod(my_series, some_integer) no longer works since version 0.13. In Pandas version 0.12, you could apply the Python built-in `divmod` function to a `Series` and an integer: ``` Enthought Canopy Python 2.7.6 \| 64-bit \| (default, Jun 4 2014, 16:42:26) [GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import pandas as pd >>> pd.__version__ '0.12.0' >>> divmod(pd.Series(range(4)), 2) (0 0 1 0 2 1 3 1 dtype: int64, 0 0 1 1 2 0 3 1 dtype: int64) >>> ``` With version >= 0.13, it appears that this usage is no longer supported: ``` Enthought Canopy Python 2.7.6 \| 64-bit \| (default, Jun 4 2014, 16:42:26) [GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import pandas as pd >>> pd.__version__ '0.14.1' >>> divmod(pd.Series(range(4)), 2) Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: unsupported operand type(s) for divmod(): 'Series' and 'int' ``` Was this change intentional? Some context: I was using this to read a climate datafile that had a 4-digit column holding combined month and day values. The original code looked something like: `month, day = divmod(df['MODA'], 100)`, but broke after upgrading to version 0.14.	`__divmod__` was not included when Series was refactored in 0.13. You can simply do: `s // 2, s % 2` if you want to get the same results If you would like to add the method, a PR would be accepted. I don't think was included as no tests for this (and to be honest never seen it used; its more of a 'functional' way of working with objects)	2016-09-12T18:58:01Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: unsupported operand type(s) for divmod(): 'Series' and 'int'	11,005
pandas-dev/pandas	pandas-dev__pandas-14225	e8357a15cd61ff698cbd3d57904133c586a8ed8b	diff --git a/doc/source/whatsnew/v0.19.0.txt b/doc/source/whatsnew/v0.19.0.txt --- a/doc/source/whatsnew/v0.19.0.txt +++ b/doc/source/whatsnew/v0.19.0.txt @@ -1572,3 +1572,4 @@ Bug Fixes - Bug in ``eval()`` where the ``resolvers`` argument would not accept a list (:issue:`14095`) - Bugs in ``stack``, ``get_dummies``, ``make_axis_dummies`` which don't preserve categorical dtypes in (multi)indexes (:issue:`13854`) - ``PeridIndex`` can now accept ``list`` and ``array`` which contains ``pd.NaT`` (:issue:`13430`) +- Bug in ``df.groupby`` where ``.median()`` returns arbitrary values if grouped dataframe contains empty bins (:issue:`13629`) diff --git a/pandas/algos.pyx b/pandas/algos.pyx --- a/pandas/algos.pyx +++ b/pandas/algos.pyx @@ -992,7 +992,7 @@ def is_lexsorted(list list_of_arrays): def groupby_indices(dict ids, ndarray[int64_t] labels, ndarray[int64_t] counts): """ - turn group_labels output into a combined indexer maping the labels to + turn group_labels output into a combined indexer mapping the labels to indexers Parameters @@ -1313,6 +1313,9 @@ cdef inline float64_t _median_linear(float64_t* a, int n): cdef float64_t result cdef float64_t* tmp + if n == 0: + return NaN + # count NAs for i in range(n): if a[i] != a[i]: diff --git a/pandas/core/groupby.py b/pandas/core/groupby.py --- a/pandas/core/groupby.py +++ b/pandas/core/groupby.py @@ -4424,12 +4424,13 @@ def _reorder_by_uniques(uniques, labels): def _groupby_indices(values): if is_categorical_dtype(values): - # we have a categorical, so we can do quite a bit # bit better than factorizing again reverse = dict(enumerate(values.categories)) codes = values.codes.astype('int64') - _, counts = _hash.value_count_int64(codes, False) + + mask = 0 <= codes + counts = np.bincount(codes[mask], minlength=values.categories.size) else: reverse, codes, counts = _algos.group_labels( _values_from_object(_ensure_object(values)))	BUG: Binned groupby median function calculates median on empty bins and outputs random numbers #### Code Sample, a copy-pastable example if possible ``` import pandas as pd d = pd.DataFrame([1,2,5,6,9,3,6,5,9,7,11,36,4,7,8,25,8,24]) b = [0,5,10,15,20,25,30,35,40,45,50,55] g = d.groupby(pd.cut(d[0],b)) print g.mean() print g.median() print g.get_group('(0, 5]').median() print g.get_group('(40, 45]').median() ``` #### Expected Output ``` 0 0 (0, 5] 3.333333 (5, 10] 7.500000 (10, 15] 11.000000 (15, 20] NaN (20, 25] 24.500000 (25, 30] NaN (30, 35] NaN (35, 40] 36.000000 (40, 45] NaN (45, 50] NaN (50, 55] NaN 0 0 (0, 5] 3.5 (5, 10] 7.5 (10, 15] 11.0 (15, 20] 18.0 (20, 25] 24.5 (25, 30] 30.5 (30, 35] 30.5 (35, 40] 36.0 (40, 45] 18.0 (45, 50] 18.0 (50, 55] 18.0 0 3.5 dtype: float64 Traceback (most recent call last): File "<ipython-input-9-0663486889da>", line 1, in <module> runfile('C:/PythonDir/test04.py', wdir='C:/PythonDir') File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 714, in runfile execfile(filename, namespace) File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 74, in execfile exec(compile(scripttext, filename, 'exec'), glob, loc) File "C:/PythonDir/test04.py", line 20, in <module> print g.get_group('(40, 45]').median() File "C:\Anaconda2\lib\site-packages\pandas\core\groupby.py", line 587, in get_group raise KeyError(name) KeyError: '(40, 45]' ``` This example shows how the median-function of the groupby object outputs a random number instead of NaN like the mean-function does when a bin is empty. Directly trying to call that bin with its key leads to an error since it doesn't exist, yet the full median output suggests it does exist and that the value might even be meaningful (like in the (15, 20] bin or the (30, 35] bin). The wrong numbers that are returned can change randomly, another possible output using the same code might look like this: ``` (0, 5] 3.500000e+00 (5, 10] 7.500000e+00 (10, 15] 1.100000e+01 (15, 20] 1.800000e+01 (20, 25] 2.450000e+01 (25, 30] 3.050000e+01 (30, 35] 3.050000e+01 (35, 40] 3.600000e+01 (40, 45] 4.927210e+165 (45, 50] 4.927210e+165 (50, 55] 4.927210e+165 ``` #### output of `pd.show_versions()` `pandas: 0.18.1`	@Khris777 Thanks for reporting! As a workaround for now, you can do: ``` In [11]: g.agg(lambda x: x.median()) Out[11]: 0 0 (0, 5] 3.5 (5, 10] 7.5 (10, 15] 11.0 (15, 20] NaN (20, 25] 24.5 (25, 30] NaN (30, 35] NaN (35, 40] 36.0 (40, 45] NaN (45, 50] NaN (50, 55] NaN ``` First time contributor, thought I'd take a look into this one. Do you think there's a more logical response than raising a KeyError to `g.get_group('(40, 45]')` ? get_group with no additional arguments is supposed to return a subset of the original dataframe with values that fall within the specified interval. If there are no values in the interval (40,45] in the original dataframe, there's no way to slice that up into a sensible response. Empty dataframe? ATM, internval types are actual string reprs (and not a distinct dtype), so yes, `g.get_group('(40, 45)')` should be a `KeyError`, just like any other indexing operation.	2016-09-15T01:57:41Z	[]	[]	Traceback (most recent call last): File "<ipython-input-9-0663486889da>", line 1, in <module> runfile('C:/PythonDir/test04.py', wdir='C:/PythonDir') File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 714, in runfile execfile(filename, namespace) File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 74, in execfile exec(compile(scripttext, filename, 'exec'), glob, loc) File "C:/PythonDir/test04.py", line 20, in <module> print g.get_group('(40, 45]').median() File "C:\Anaconda2\lib\site-packages\pandas\core\groupby.py", line 587, in get_group raise KeyError(name) KeyError: '(40, 45]'	11,007
pandas-dev/pandas	pandas-dev__pandas-14329	6dcc23862b6b60ce2a67436b4a278fbe4c05490f	diff --git a/pandas/core/frame.py b/pandas/core/frame.py --- a/pandas/core/frame.py +++ b/pandas/core/frame.py @@ -1635,7 +1635,8 @@ def to_latex(self, buf=None, columns=None, col_space=None, header=True, When set to False prevents from escaping latex special characters in column names. encoding : str, default None - Default encoding is ascii in Python 2 and utf-8 in Python 3 + A string representing the encoding to use in the output file, + defaults to 'ascii' on Python 2 and 'utf-8' on Python 3. decimal : string, default '.' Character recognized as decimal separator, e.g. ',' in Europe diff --git a/pandas/formats/format.py b/pandas/formats/format.py --- a/pandas/formats/format.py +++ b/pandas/formats/format.py @@ -654,6 +654,9 @@ def to_latex(self, column_format=None, longtable=False, encoding=None): latex_renderer = LatexFormatter(self, column_format=column_format, longtable=longtable) + if encoding is None: + encoding = 'ascii' if compat.PY2 else 'utf-8' + if hasattr(self.buf, 'write'): latex_renderer.write_result(self.buf) elif isinstance(self.buf, compat.string_types):	TST: 3.5 c-locale https://travis-ci.org/pydata/pandas/jobs/161159736 xref #14114, #12337 ``` ====================================================================== ERROR: test_to_latex_filename (pandas.tests.formats.test_format.TestDataFrameFormatting) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/formats/test_format.py", line 2825, in test_to_latex_filename df.to_latex(path) File "/home/travis/build/pydata/pandas/pandas/core/frame.py", line 1661, in to_latex encoding=encoding) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 662, in to_latex latex_renderer.write_result(f) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 906, in write_result buf.write(' & '.join(crow)) UnicodeEncodeError: 'ascii' codec can't encode character '\xdf' in position 7: ordinal not in range(128) ---------------------------------------------------------------------- Ran 10273 tests in 375.734s ``` cc @nbonnotte	cc @nbonnotte	2016-10-01T17:34:21Z	[]	[]	Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/formats/test_format.py", line 2825, in test_to_latex_filename df.to_latex(path) File "/home/travis/build/pydata/pandas/pandas/core/frame.py", line 1661, in to_latex encoding=encoding) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 662, in to_latex latex_renderer.write_result(f) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 906, in write_result buf.write(' & '.join(crow)) UnicodeEncodeError: 'ascii' codec can't encode character '\xdf' in position 7: ordinal not in range(128)	11,016
pandas-dev/pandas	pandas-dev__pandas-14853	b6de920d8c3c2becc46b4fe233e9f388947554f2		BUG: Passing ambiguous ndarray[datetime64[ns]] to DatetimeIndex constructor can cause ValueError with wrong offset if you give infer_freq 5 consecutive weekdays, it'll come back with 'D' as its inferred frequency. But if your actual frequency is `BDay`, then, when DatetimeIndex checks that the frequency matches, 'B' != 'D'. (note that verify_integrity=False skips this). This leads to a more general issue about `infer_freq` with ambiguous cases. I think it makes the most sense to move these sorts of checks to a method on offset that takes a frequence and an Index or ndarray, and determines whether it is compatible. This matters because you can hit some edge cases when you pass freq and also datetime64[ns] to the DatetimeIndex constructor and more generally because comparing freqstr is probably not the best way to go about checking whether a frequency matches. Default implementation could be: ``` python def is_compatible(freqstr, arr=None): return freqstr == self.freqstr ``` and then bday could do something like (and this is totally psuedocode) ``` python def is_compatible(freqstr, arr=None): if freqstr == self.freqstr: return True if arr is not None and len(arr) <= 5: if freqstr == 'D': # or other compatibles that ensure it's consecutive return all(is_weekday(date) for date in arr) ``` This gets more complicated with multiplied offsets, but I think it's worth considering. ``` python from datetime import datetime import pandas as pd dates = [datetime(2013, 10, 7), datetime(2013, 10, 8), datetime(2013, 10, 9)] ind = pd.DatetimeIndex(dates, freq=pd.tseries.frequencies.BDay()) ind2 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay(), verify_integrity=False) ind3 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay()) ``` produces this Traceback: ``` Traceback (most recent call last): File "test2.py", line 8, in <module> ind3 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay()) File "../pandas/tseries/index.py", line 280, in __new__ raise ValueError('Dates do not conform to passed ' ValueError: Dates do not conform to passed frequency ``` cc @cancan101 - this is what we need to deal with in adding your offsets. I believe that every other offset can be returned from infer_freq, so these offsets would be different and therefore could _never_ pass integrity checks. So either we'd need to change infer_freq and/or define some kind of is_compatible method that intelligently covers all the ways in which the frequency could be something different than its freqstr.	Pretty simple test case: ``` python In [88]: infer_freq([datetime(2013, 10, 7), datetime(2013, 10, 8), datetime(2013, 10, 9)]) Out[88]: 'D' ``` which I agree is incorrect since a frequency of 'BD' cannot be ruled out. @wesm ? no that's not incorrect - it's reasonable and valid as the freq for the sequence, right? It might be ambiguous, but all that matters is inferring some frequency. So either it should return a list of possibilities, or we pass the buck to offsets to handle the ambiguity. Okay. Agreed, "incorrect" is not the right word. That being said, what exactly is the spec for infer_freq? There is not quite a total ordering of frequencies by specificity, but in general, should infer_freq return the most specific or the more general frequency? @cancan101 it obviously is the most general, it is used quite extensively interally to lazily evaluate frequency when its not already assigned Perhaps `infer_freq` should have an option not to guess when there is any potential for ambiguity. This might be the default behavior, then. I don't think anyone will shed tears if a length-3 (or 5) array case like you describe gets inferred as no frequency with the change. @wesm or at the very least could skip frequency inference when small.	2016-12-10T21:23:26Z	[]	[]	Traceback (most recent call last): File "test2.py", line 8, in <module> ind3 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay()) File "../pandas/tseries/index.py", line 280, in __new__ raise ValueError('Dates do not conform to passed ' ValueError: Dates do not conform to passed frequency	11,066
pandas-dev/pandas	pandas-dev__pandas-14884	3ba2cff9c55cd16b172f9feb09da551990753f3b		Assigning datetime array to column fails with OutOfBoundsDatetime when having NaT and other unit as [ns] Assigning an array with datetime64[ns] values including a NaT just works: ``` In [85]: a = np.array([1, 'nat'], dtype='datetime64[ns]') In [86]: pd.Series(a) Out[86]: 0 1970-01-01 00:00:00.000000001 1 NaT dtype: datetime64[ns] In [88]: df = pd.Series(a).to_frame() In [89]: df['new'] = a ``` But when having an array with another date unit, converting it to a Series still works, but assigning it directly to a column not anymore, resulting in a OutOfBoundsDatetime error: ``` In [90]: a = np.array([1, 'nat'], dtype='datetime64[s]') In [91]: pd.Series(a) Out[91]: 0 1970-01-01 00:00:01 1 NaT dtype: datetime64[ns] In [92]: df['new'] = a Traceback (most recent call last): ... File "tslib.pyx", line 1720, in pandas.tslib.cast_to_nanoseconds (pandas\tslib.c:27435) File "tslib.pyx", line 1023, in pandas.tslib._check_dts_bounds (pandas\tslib.c:18102) OutOfBoundsDatetime: Out of bounds nanosecond timestamp: 292277026596-12-03 08:29:52 ``` If you first convert it to a series, it does work. Also if the `NaT` is not present: ``` In [93]: df['new'] = pd.Series(a) In [94]: a = np.array([1, 2], dtype='datetime64[s]') In [95]: df['new'] = a ```	yep, this conversion is amazingly complex. marking as a bug	2016-12-15T06:06:56Z	[]	[]	Traceback (most recent call last): ... File "tslib.pyx", line 1720, in pandas.tslib.cast_to_nanoseconds (pandas\tslib.c:27435) File "tslib.pyx", line 1023, in pandas.tslib._check_dts_bounds (pandas\tslib.c:18102) OutOfBoundsDatetime: Out of bounds nanosecond timestamp: 292277026596-12-03 08:29:52	11,069
pandas-dev/pandas	pandas-dev__pandas-14886	5f889a2106f6584583458e01dbd0f3b9b696fab2	diff --git a/doc/source/whatsnew/v0.19.2.txt b/doc/source/whatsnew/v0.19.2.txt --- a/doc/source/whatsnew/v0.19.2.txt +++ b/doc/source/whatsnew/v0.19.2.txt @@ -78,7 +78,7 @@ Bug Fixes - Bug in clipboard functions on linux with python2 with unicode and separators (:issue:`13747`) - Bug in clipboard functions on Windows 10 and python 3 (:issue:`14362`, :issue:`12807`) - Bug in ``.to_clipboard()`` and Excel compat (:issue:`12529`) - +- Bug in ``DataFrame.combine_first()`` for integer columns (:issue:`14687`). - Bug in ``pd.read_csv()`` in which the ``dtype`` parameter was not being respected for empty data (:issue:`14712`) - Bug in ``pd.read_csv()`` in which the ``nrows`` parameter was not being respected for large input when using the C engine for parsing (:issue:`7626`) diff --git a/pandas/core/frame.py b/pandas/core/frame.py --- a/pandas/core/frame.py +++ b/pandas/core/frame.py @@ -3665,10 +3665,8 @@ def combine(self, other, func, fill_value=None, overwrite=True): otherSeries[other_mask] = fill_value # if we have different dtypes, possibily promote - if notnull(series).all(): - new_dtype = this_dtype - otherSeries = otherSeries.astype(new_dtype) - else: + new_dtype = this_dtype + if not is_dtype_equal(this_dtype, other_dtype): new_dtype = _find_common_type([this_dtype, other_dtype]) if not is_dtype_equal(this_dtype, new_dtype): series = series.astype(new_dtype)	combine_first throws ValueError: Cannot convert NA to integer I do not understand why there is a need to convert NA to integer if the result does not have NAs. Perhaps the combine_first algo needs to do it under the hood? #### A small, complete example of the issue ```python from pandas import DataFrame DataFrame({'a': [0, 1, 3, 5]}).combine_first(DataFrame({'a': [1, 4]})) Traceback (most recent call last): File "/usr/local/lib/python3.4/dist-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-16-12b973b1b150>", line 1, in <module> pd.DataFrame({'a': [0, 1, 3, 5]}).combine_first(pd.DataFrame({'a': [1, 4]})) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3787, in combine_first return self.combine(other, combiner, overwrite=False) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3714, in combine otherSeries = otherSeries.astype(new_dtype) File "/usr/local/lib/python3.4/dist-packages/pandas/core/generic.py", line 3054, in astype raise_on_error=raise_on_error, kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3168, in astype return self.apply('astype', dtype=dtype, kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3035, in apply applied = getattr(b, f)(kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 462, in astype values=values, kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 505, in _astype values = _astype_nansafe(values.ravel(), dtype, copy=True) File "/usr/local/lib/python3.4/dist-packages/pandas/types/cast.py", line 531, in _astype_nansafe raise ValueError('Cannot convert NA to integer') ValueError: Cannot convert NA to integer ``` #### Expected Output ``` a 0 0 1 1 2 3 3 5 ``` It does work when at least one item is a float: ```python DataFrame({'a': [0.0, 1, 3, 5]}).combine_first(DataFrame({'a': [1, 4]})) a 0 0.0 1 1.0 2 3.0 3 5.0 ``` I am aware that integer series cannot have NAs but there is no need to introduce NAs here. I do like it that the series is not upcasted to float silently though. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.4.3.final.0 python-bits: 64 OS: Linux OS-release: 3.19.0-66-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.19.0 nose: None pip: 1.5.4 setuptools: 3.3 Cython: 0.24.1 numpy: 1.11.2 scipy: 0.17.1 statsmodels: 0.6.1 xarray: None IPython: 4.0.0 sphinx: None patsy: 0.4.1 dateutil: 2.5.3 pytz: 2016.7 blosc: None bottleneck: None tables: 3.2.2 numexpr: 2.4.6 matplotlib: 1.5.1 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: 0.999 httplib2: 0.9.2 apiclient: None sqlalchemy: None pymysql: None psycopg2: 2.6.1 (dt dec pq3 ext lo64) jinja2: 2.8 boto: None pandas_datareader: None </details>	This seems to be a regression from 0.18, as this worked before: ``` In [1]: DataFrame({'a': [0, 1, 3, 5]}).combine_first(DataFrame({'a': [1, 4]})) Out[1]: a 0 0 1 1 2 3 3 5 In [2]: pd.__version__ Out[2]: u'0.18.1' ``` @Dmitrii-I Thanks for the report! Always welcome to look into what could have caused this change.	2016-12-15T10:53:24Z	[]	[]	Traceback (most recent call last): File "/usr/local/lib/python3.4/dist-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-16-12b973b1b150>", line 1, in <module> pd.DataFrame({'a': [0, 1, 3, 5]}).combine_first(pd.DataFrame({'a': [1, 4]})) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3787, in combine_first return self.combine(other, combiner, overwrite=False) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3714, in combine otherSeries = otherSeries.astype(new_dtype) File "/usr/local/lib/python3.4/dist-packages/pandas/core/generic.py", line 3054, in astype raise_on_error=raise_on_error, kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3168, in astype return self.apply('astype', dtype=dtype, kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3035, in apply applied = getattr(b, f)(kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 462, in astype values=values, kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 505, in _astype values = _astype_nansafe(values.ravel(), dtype, copy=True) File "/usr/local/lib/python3.4/dist-packages/pandas/types/cast.py", line 531, in _astype_nansafe raise ValueError('Cannot convert NA to integer') ValueError: Cannot convert NA to integer	11,070
pandas-dev/pandas	pandas-dev__pandas-14907	e503d40ace473556990e5453ed5b4c9aa96e24ff		groupby/transform with NaNs in grouped column What's the expected behavior when grouping on a column containing `NaN` and then applying `transform`? For a `Series`, the current result is to throw an exception: ``` >>> df = pd.DataFrame({ ... 'a' : range(10), ... 'b' : [1, 1, 2, 3, np.nan, 4, 4, 5, 5, 5]}) >>> >>> df.groupby(df.b)['a'].transform(max) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/groupby.py", line 2422, in transform return self._transform_fast(cyfunc) File "pandas/core/groupby.py", line 2463, in _transform_fast return self._set_result_index_ordered(Series(values)) File "pandas/core/groupby.py", line 498, in _set_result_index_ordered result.index = self.obj.index File "pandas/core/generic.py", line 1997, in __setattr__ return object.__setattr__(self, name, value) File "pandas/src/properties.pyx", line 65, in pandas.lib.AxisProperty.__set__ (pandas/lib.c:41301) obj._set_axis(self.axis, value) File "pandas/core/series.py", line 273, in _set_axis self._data.set_axis(axis, labels) File "pandas/core/internals.py", line 2219, in set_axis 'new values have %d elements' % (old_len, new_len)) ValueError: Length mismatch: Expected axis has 9 elements, new values have 10 elements ``` For a `DataFrame`, the missing value gets filled in with what looks like an uninitialized value from `np.empty_like`: ``` >>> df.groupby(df.b).transform(max) a 0 1 1 1 2 2 3 3 4 -1 5 6 6 6 7 9 8 9 9 9 ``` It seems like either it should fill in the missing values with `NaN` (which might require a change of dtype), or just drop those rows from the result (which requires the shape to change). Either solution has the potential to surprise.	http://pandas.pydata.org/pandas-docs/stable/groupby.html#na-group-handling This _should_ work, so this is a bug as the NA group is not defined. Resultant value should be `NaN`. xref #5456 xref #6992 xref #443	2016-12-18T07:12:07Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/groupby.py", line 2422, in transform return self._transform_fast(cyfunc) File "pandas/core/groupby.py", line 2463, in _transform_fast return self._set_result_index_ordered(Series(values)) File "pandas/core/groupby.py", line 498, in _set_result_index_ordered result.index = self.obj.index File "pandas/core/generic.py", line 1997, in __setattr__ return object.__setattr__(self, name, value) File "pandas/src/properties.pyx", line 65, in pandas.lib.AxisProperty.__set__ (pandas/lib.c:41301) obj._set_axis(self.axis, value) File "pandas/core/series.py", line 273, in _set_axis self._data.set_axis(axis, labels) File "pandas/core/internals.py", line 2219, in set_axis 'new values have %d elements' % (old_len, new_len)) ValueError: Length mismatch: Expected axis has 9 elements, new values have 10 elements	11,072
pandas-dev/pandas	pandas-dev__pandas-14952	74de478392e09cf938d244f5990da4e001afc84c		Groupby.groups doesn't work by a groups convert from DateTimeIndex ``` i = pd.DatetimeIndex(pd.date_range('2015/01/01', periods=5), name='date') d = pd.DataFrame({'A':[5,6,7,8,9], 'B':[1,2,3,4,5]}, index=i) print i print d dg = d.groupby(level='date') print dg.get_group('2015-01-01') print dg.groups ``` Output: ``` DatetimeIndex(['2015-01-01', '2015-01-02', '2015-01-03', '2015-01-04', '2015-01-05'], dtype='datetime64[ns]', name=u'date', freq='D', tz=None) A B date 2015-01-01 5 1 2015-01-02 6 2 2015-01-03 7 3 2015-01-04 8 4 2015-01-05 9 5 A B date 2015-01-01 5 1 Traceback (most recent call last): File "<ipython-input-2-f2b4d1146750>", line 12, in <module> print dg.groups File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 405, in groups return self.grouper.groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 1352, in groups return self.groupings[0].groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 2041, in groups return self.index.groupby(self.grouper) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/tseries/base.py", line 60, in groupby return _algos.groupby_object(objs, f) TypeError: Argument 'labels' has incorrect type (expected numpy.ndarray, got DatetimeIndex) ```	hmm, does seem a bit buggy @Temmplar reason you closed this? @jreback sorry I guess, I'm not familiar with the system and I thought if you answered I have to close it. nope would be closed by a pull request to fix which are welcome! I tried this in `tseries/base.py`: ``` - return _algos.groupby_object(objs, f) + return _algos.groupby_object(objs, np.asarray(f)) ``` And it sort of works, but the results are ugly, especially if a timezone is involved. An index like this: ``` DatetimeIndex(['2016-06-28 05:30:00-05:00', '2016-06-28 05:31:00-05:00'], dtype='datetime64[ns, America/Chicago]') ``` Produces naive UTC results: ``` {numpy.datetime64('2016-06-28T10:30:00.000000000'): [Timestamp('2016-06-28 05:30:00-0500', ... ``` @jreback Do you have any idea how to fix this simply and properly? I am getting this error too.	2016-12-22T06:45:33Z	[]	[]	Traceback (most recent call last): File "<ipython-input-2-f2b4d1146750>", line 12, in <module> print dg.groups File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 405, in groups return self.grouper.groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 1352, in groups return self.groupings[0].groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 2041, in groups return self.index.groupby(self.grouper) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/tseries/base.py", line 60, in groupby return _algos.groupby_object(objs, f) TypeError: Argument 'labels' has incorrect type (expected numpy.ndarray, got DatetimeIndex)	11,079
pandas-dev/pandas	pandas-dev__pandas-15569	5f0b69aee3622eed9392cef163e4b31ba742498e	diff --git a/doc/source/whatsnew/v0.20.0.txt b/doc/source/whatsnew/v0.20.0.txt --- a/doc/source/whatsnew/v0.20.0.txt +++ b/doc/source/whatsnew/v0.20.0.txt @@ -635,7 +635,7 @@ Performance Improvements - Increased performance of ``pd.factorize()`` by releasing the GIL with ``object`` dtype when inferred as strings (:issue:`14859`) - Improved performance of timeseries plotting with an irregular DatetimeIndex (or with ``compat_x=True``) (:issue:`15073`). -- Improved performance of ``groupby().cummin()`` and ``groupby().cummax()`` (:issue:`15048`, :issue:`15109`) +- Improved performance of ``groupby().cummin()`` and ``groupby().cummax()`` (:issue:`15048`, :issue:`15109`, :issue:`15561`) - Improved performance and reduced memory when indexing with a ``MultiIndex`` (:issue:`15245`) - When reading buffer object in ``read_sas()`` method without specified format, filepath string is inferred rather than buffer object. (:issue:`14947`) - Improved performance of `rank()` for categorical data (:issue:`15498`) diff --git a/pandas/core/groupby.py b/pandas/core/groupby.py --- a/pandas/core/groupby.py +++ b/pandas/core/groupby.py @@ -1442,7 +1442,7 @@ def cummin(self, axis=0, kwargs): if axis != 0: return self.apply(lambda x: np.minimum.accumulate(x, axis)) - return self._cython_transform('cummin', kwargs) + return self._cython_transform('cummin', numeric_only=False) @Substitution(name='groupby') @Appender(_doc_template) @@ -1451,7 +1451,7 @@ def cummax(self, axis=0, kwargs): if axis != 0: return self.apply(lambda x: np.maximum.accumulate(x, axis)) - return self._cython_transform('cummax', kwargs) + return self._cython_transform('cummax', numeric_only=False) @Substitution(name='groupby') @Appender(_doc_template)	BUG: cython version of groupby.cummax throws error on datetimes #### Code Sample, a copy-pastable example if possible ```python >>> import pandas as pd >>> x = pd.DataFrame(dict(a=[1], b=pd.to_datetime(['2001']))) >>> x.groupby('a').b.cummax() Traceback (most recent call last): File "<ipython-input-9-316257648d5f>", line 1, in <module> x.groupby('a').b.cummax() File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1454, in cummax return self._cython_transform('cummax', **kwargs) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 806, in _cython_transform raise DataError('No numeric types to aggregate') DataError: No numeric types to aggregate ``` #### Problem description The current github version of pandas has cython implementations of `groupby.cummin` and `groupby.cummax`, which throw an error if called on datetime columns. (See #15048, 0fe491d.) #### Expected Output ```python 0 2001-01-01 Name: b, dtype: datetime64[ns] ``` #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.5.2.final.0 python-bits: 64 OS: Linux OS-release: 4.9.8-100.fc24.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: C LANG: C LOCALE: None.None pandas: 0.19.0+531.g04e1168 pytest: 3.0.5 pip: 9.0.1 setuptools: 27.2.0 Cython: 0.25.2 numpy: 1.11.3 scipy: 0.18.1 xarray: 0.9.1 IPython: 4.2.0 sphinx: 1.5.1 patsy: 0.4.1 dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.2.0 tables: 3.3.0 numexpr: 2.6.2 feather: None matplotlib: 2.0.0 openpyxl: 2.4.1 xlrd: 1.0.0 xlwt: 1.2.0 xlsxwriter: 0.9.6 lxml: 3.7.2 bs4: 4.5.3 html5lib: 0.999 sqlalchemy: 1.1.5 pymysql: None psycopg2: None jinja2: 2.9.4 s3fs: None pandas_gbq: None pandas_datareader: None </details>	I think this may be as easy as adding a `numeric_only=False` to this function call. (of course with tests) https://github.com/pandas-dev/pandas/blob/04e116851337cd852b4255f8221d9be44829e0e1/pandas/core/groupby.py#L1454 This is correct and not a bug. datetimes are not numeric types. This has to be specifically enabled. this was covered by #15054 ``` In [29]: x = pd.DataFrame(dict(a=[1], b=pd.to_datetime(['2001']))) ...: >>> x.groupby('a').b.cummax(numeric_only=False) ...: ...: Out[29]: 0 2001-01-01 Name: b, dtype: datetime64[ns] ``` @jreback - maybe the API is set, but isn't this inconsistent? ```python In [23]: df = pd.DataFrame(dict(a=[1], b=pd.to_datetime(['2001']))) In [24]: df['b'].max() Out[24]: Timestamp('2001-01-01 00:00:00') In [25]: df['b'].cummax() Out[25]: 0 2001-01-01 Name: b, dtype: datetime64[ns] In [26]: df.groupby('a')['b'].max() Out[26]: a 1 2001-01-01 Name: b, dtype: datetime64[ns] In [27]: df.groupby('a')['b'].cummax() Out[27]: DataError: No numeric types to aggregate ``` @chris-b1 hmm, ahh we are defaulting ``numeric_only=False`` for ``.max`` yes that should be done for ``cummax``/``cummin``.	2017-03-04T20:20:15Z	[]	[]	Traceback (most recent call last): File "<ipython-input-9-316257648d5f>", line 1, in <module> x.groupby('a').b.cummax() File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1454, in cummax return self._cython_transform('cummax', **kwargs) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 806, in _cython_transform raise DataError('No numeric types to aggregate') DataError: No numeric types to aggregate	11,116
pandas-dev/pandas	pandas-dev__pandas-16090	d313e4dd7605a658869f5d026d6705afb169ab40	diff --git a/doc/source/whatsnew/v0.20.0.txt b/doc/source/whatsnew/v0.20.0.txt --- a/doc/source/whatsnew/v0.20.0.txt +++ b/doc/source/whatsnew/v0.20.0.txt @@ -1600,7 +1600,7 @@ Indexing - Bug in the HTML display with with a ``MultiIndex`` and truncation (:issue:`14882`) - Bug in the display of ``.info()`` where a qualifier (+) would always be displayed with a ``MultiIndex`` that contains only non-strings (:issue:`15245`) - Bug in ``pd.concat()`` where the names of ``MultiIndex`` of resulting ``DataFrame`` are not handled correctly when ``None`` is presented in the names of ``MultiIndex`` of input ``DataFrame`` (:issue:`15787`) -- Bug in ``DataFrame.sort_index()`` and ``Series.sort_index()`` where ``na_position`` doesn't work with a ``MultiIndex`` (:issue:`14784`) +- Bug in ``DataFrame.sort_index()`` and ``Series.sort_index()`` where ``na_position`` doesn't work with a ``MultiIndex`` (:issue:`14784`, :issue:`16604`) I/O ^^^ diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py --- a/pandas/core/indexes/multi.py +++ b/pandas/core/indexes/multi.py @@ -1645,10 +1645,11 @@ def _get_labels_for_sorting(self): """ from pandas.core.categorical import Categorical - return [Categorical.from_codes(label, - np.arange(np.array(label).max() + 1, - dtype=label.dtype), - ordered=True) + def cats(label): + return np.arange(np.array(label).max() + 1 if len(label) else 0, + dtype=label.dtype) + + return [Categorical.from_codes(label, cats(label), ordered=True) for label in self.labels] def sortlevel(self, level=0, ascending=True, sort_remaining=True):	BUG: groupy().nth() throws error on multiple groups, empty result #### Code Sample, a copy-pastable example if possible ```python >>> import pandas as pd >>> df = pd.DataFrame(index=[0], columns=['a', 'b', 'c']) >>> df.groupby('a').nth(10) Empty DataFrame Columns: [b, c] Index: [] >>> df.groupby(['a', 'b']).nth(10) Traceback (most recent call last): File "<ipython-input-3-ae8299c3984e>", line 1, in <module> df.groupby(['a', 'b']).nth(10) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1390, in nth return out.sort_index() if self.sort else out File "~/anaconda3/lib/python3.5/site-packages/pandas/core/frame.py", line 3344, in sort_index indexer = lexsort_indexer(labels._get_labels_for_sorting(), File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in _get_labels_for_sorting for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in <listcomp> for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/numpy/core/_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) ValueError: zero-size array to reduction operation maximum which has no identity ``` #### Problem description In the current Github version of Pandas, when calling `groupby().nth()` with multiple grouping columns, an error is raised if the result is empty. This is a regression from version 0.19.2. #### Expected Output ```python Empty DataFrame Columns: [b, c] Index: [] Empty DataFrame Columns: [c] Index: [] ``` #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.5.2.final.0 python-bits: 64 OS: Linux OS-release: 4.9.8-100.fc24.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: C LANG: C LOCALE: None.None pandas: 0.19.0+829.gb17e286 pytest: 3.0.5 pip: 9.0.1 setuptools: 27.2.0 Cython: 0.25.2 numpy: 1.11.3 scipy: 0.18.1 xarray: 0.9.1 IPython: 4.2.0 sphinx: 1.5.1 patsy: 0.4.1 dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.2.0 tables: 3.3.0 numexpr: 2.6.2 feather: None matplotlib: 2.0.0 openpyxl: 2.4.1 xlrd: 1.0.0 xlwt: 1.2.0 xlsxwriter: 0.9.6 lxml: 3.7.2 bs4: 4.5.3 html5lib: 0.999 sqlalchemy: 1.1.5 pymysql: None psycopg2: None jinja2: 2.9.4 s3fs: None pandas_gbq: None pandas_datareader: None </details>	Think this just requires: ```diff diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py index 92baf9d..34b62c5 100644 --- a/pandas/core/indexes/multi.py +++ b/pandas/core/indexes/multi.py @@ -1645,11 +1645,9 @@ class MultiIndex(Index): """ from pandas.core.categorical import Categorical - return [Categorical.from_codes(label, - np.arange(np.array(label).max() + 1, - dtype=label.dtype), - ordered=True) - for label in self.labels] + return [Categorical.from_codes(label, np.arange( + np.array(label).max() + 1 if len(label) else 0, + dtype=label.dtype), ordered=True) for label in self.labels] def sortlevel(self, level=0, ascending=True, sort_remaining=True): """ ``` can u put up a PR with that fix (and test)?	2017-04-21T23:11:51Z	[]	[]	Traceback (most recent call last): File "<ipython-input-3-ae8299c3984e>", line 1, in <module> df.groupby(['a', 'b']).nth(10) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1390, in nth return out.sort_index() if self.sort else out File "~/anaconda3/lib/python3.5/site-packages/pandas/core/frame.py", line 3344, in sort_index indexer = lexsort_indexer(labels._get_labels_for_sorting(), File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in _get_labels_for_sorting for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in <listcomp> for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/numpy/core/_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) ValueError: zero-size array to reduction operation maximum which has no identity	11,152
pandas-dev/pandas	pandas-dev__pandas-16294	4bed864a24901d9c2baab5e17c57c956a188602f	diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -48,7 +48,7 @@ Indexing I/O ^^^ - +- Bug that would force importing of the clipboard routines unecessarily, potentially causing an import error on startup (:issue:`16288`) Plotting diff --git a/pandas/core/generic.py b/pandas/core/generic.py --- a/pandas/core/generic.py +++ b/pandas/core/generic.py @@ -1382,8 +1382,8 @@ def to_clipboard(self, excel=None, sep=None, kwargs): - Windows: none - OS X: none """ - from pandas.io.clipboard import clipboard - clipboard.to_clipboard(self, excel=excel, sep=sep, kwargs) + from pandas.io import clipboards + clipboards.to_clipboard(self, excel=excel, sep=sep, **kwargs) def to_xarray(self): """ diff --git a/pandas/io/api.py b/pandas/io/api.py --- a/pandas/io/api.py +++ b/pandas/io/api.py @@ -5,7 +5,7 @@ # flake8: noqa from pandas.io.parsers import read_csv, read_table, read_fwf -from pandas.io.clipboard.clipboard import read_clipboard +from pandas.io.clipboards import read_clipboard from pandas.io.excel import ExcelFile, ExcelWriter, read_excel from pandas.io.pytables import HDFStore, get_store, read_hdf from pandas.io.json import read_json diff --git a/pandas/io/clipboard/clipboard.py b/pandas/io/clipboards.py similarity index 100% rename from pandas/io/clipboard/clipboard.py rename to pandas/io/clipboards.py	ImportError with pandas 0.20.0 and 0.20.1 #### Code Sample, a copy-pastable example if possible With versions 0.20.0 and 0.20.1, I get the following error: ```python import pandas ``` ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/__init__.py", line 58, in <module> from pandas.io.api import * File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/api.py", line 8, in <module> from pandas.io.clipboard.clipboard import read_clipboard File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 103, in <module> copy, paste = determine_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 76, in determine_clipboard return init_qt_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/clipboards.py", line 49, in init_qt_clipboard from PyQt4.QtGui import QApplication ImportError: No module named sip ``` This does not occur with 0.19.2 Here are the other packages installed in my virtualenv ``` $ pip freeze appdirs==1.4.3 numpy==1.12.1 packaging==16.8 pandas==0.20.1 pyparsing==2.2.0 python-dateutil==2.6.0 pytz==2017.2 PyYAML==3.12 six==1.10.0 ``` #### Problem description It would be nice if pandas was pip installable. As it stands, the new versions are not (for me). #### Expected Output No error when importing pandas. #### Output of ``pd.show_versions()`` I can't get that info without being able to import pandas.	@jorisvandenbossche did you see something like this earlier? Not exactly this error, but possibly related. For me the error with clipboard was something with ``from PyQt4 import QtCore``. For some reason, I had an empty PyQt4 package in site-packages next to PyQt5 (so `import PyQt4` did work, which is used to check which clipboard backend to use, but then later on actual imports raise an error), and for some reason this started to give problems. Matplotlib also had this problem, so I am not sure it was pandas-related. @FragLegs Is this in a clean (newly made) env? Or if not, can you see if you can reproduce it then as well?	2017-05-09T01:07:17Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/__init__.py", line 58, in <module> from pandas.io.api import * File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/api.py", line 8, in <module> from pandas.io.clipboard.clipboard import read_clipboard File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 103, in <module> copy, paste = determine_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 76, in determine_clipboard return init_qt_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/clipboards.py", line 49, in init_qt_clipboard from PyQt4.QtGui import QApplication ImportError: No module named sip	11,183
pandas-dev/pandas	pandas-dev__pandas-16434	49ec31bbaeca81a6f58fc1be26fe80f3ac188cdd	diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -80,7 +80,7 @@ Reshaping ^^^^^^^^^ - Bug in ``DataFrame.stack`` with unsorted levels in MultiIndex columns (:issue:`16323`) - +- Bug in ``Series.isin(..)`` with a list of tuples (:issue:`16394`) Numeric ^^^^^^^ diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -388,7 +388,7 @@ def isin(comps, values): "[{0}]".format(type(values).__name__)) if not isinstance(values, (ABCIndex, ABCSeries, np.ndarray)): - values = np.array(list(values), dtype='object') + values = lib.list_to_object_array(list(values)) comps, dtype, _ = _ensure_data(comps) values, _, _ = _ensure_data(values, dtype=dtype)	BUG: ValueError with Series.isin and tuples #### Code Sample, a copy-pastable example if possible ``` import pandas as pd df = pd.DataFrame({'A': [1, 2, 3], 'B': ['a', 'b', 'f']}) df['C'] = list(zip(df['A'], df['B'])) df['C'].isin([(1, 'a')]) ``` #### Problem description Returns ValueError: Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/series.py", line 2555, in isin result = algorithms.isin(_values_from_object(self), values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 421, in isin return f(comps, values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 399, in <lambda> f = lambda x, y: htable.ismember_object(x, values) File "pandas/_libs/hashtable_func_helper.pxi", line 428, in pandas._libs.hashtable.ismember_object (pandas/_libs/hashtable.c:29677) ValueError: Buffer has wrong number of dimensions (expected 1, got 2) #### Expected Output In pandas 0.19.2 returns: 0 True 1 False 2 False Name: C, dtype: bool #### Output of ``pd.show_versions()`` <details> # Paste the output here pd.show_versions() here INSTALLED VERSIONS ------------------ commit: None python: 3.6.1.final.0 python-bits: 64 OS: Darwin OS-release: 16.5.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.20.0rc2 pytest: None pip: 9.0.1 setuptools: 27.2.0 Cython: None numpy: 1.12.1 scipy: None xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None </details>	this code was refactored to be more general, so this was a missing case. easy fix I think. ``np.array`` converts nested tuples to lists, which is not nice, so do this. if you'd like to submit a PR with this as an added tests (and make sure nothing else breaks), would be great. ``` diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py index a745ec6..77d79c9 100644 --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -388,7 +388,7 @@ def isin(comps, values): "[{0}]".format(type(values).__name__)) if not isinstance(values, (ABCIndex, ABCSeries, np.ndarray)): - values = np.array(list(values), dtype='object') + values = lib.list_to_object_array(list(values)) comps, dtype, _ = _ensure_data(comps) values, _, _ = _ensure_data(values, dtype=dtype) ``` I'm taking a crack at this. Is the solution to just add lib.list_to_object_array back in along with a test for the tuple case, or should we check if comps contains tuples and use lib.list_to_object_array only if it does?	2017-05-22T20:03:21Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/series.py", line 2555, in isin result = algorithms.isin(_values_from_object(self), values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 421, in isin return f(comps, values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 399, in <lambda> f = lambda x, y: htable.ismember_object(x, values) File "pandas/_libs/hashtable_func_helper.pxi", line 428, in pandas._libs.hashtable.ismember_object (pandas/_libs/hashtable.c:29677) ValueError: Buffer has wrong number of dimensions (expected 1, got 2)	11,206
pandas-dev/pandas	pandas-dev__pandas-16486	92d07992e826808cd56f0bd8fec083b510ca402d	diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -40,6 +40,7 @@ Bug Fixes - Silenced a warning on some Windows environments about "tput: terminal attributes: No such device or address" when detecting the terminal size. This fix only applies to python 3 (:issue:`16496`) - Bug in using ``pathlib.Path`` or ``py.path.local`` objects with io functions (:issue:`16291`) +- Bug in ``Index.symmetric_difference()`` on two equal MultiIndex's, results in a TypeError (:issue `13490`) - Bug in ``DataFrame.update()`` with ``overwrite=False`` and ``NaN values`` (:issue:`15593`) diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py --- a/pandas/core/indexes/multi.py +++ b/pandas/core/indexes/multi.py @@ -414,6 +414,12 @@ def view(self, cls=None): return result def _shallow_copy_with_infer(self, values=None, kwargs): + # On equal MultiIndexes the difference is empty. + # Therefore, an empty MultiIndex is returned GH13490 + if len(values) == 0: + return MultiIndex(levels=[[] for _ in range(self.nlevels)], + labels=[[] for _ in range(self.nlevels)], + kwargs) return self._shallow_copy(values, **kwargs) @Appender(_index_shared_docs['_shallow_copy'])	Symmetric difference on equal MultiIndexes raises TypeError Calling `symmetric_difference` on two equal multiindices results in a TypeError rather than an empty MultiIndex. This is surprising since calling `difference` on the same multiindices results in the expected empty MultiIndex. #### Code Sample, a copy-pastable example if possible ``` a = pandas.MultiIndex.from_product([['a', 'b'], [0, 1]]) b = pandas.MultiIndex.from_product([['a', 'b'], [0, 1]]) print(a.symmetric_difference(b)) ``` Which gives the following error: ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1674, in __xor__ return self.symmetric_difference(other) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1911, in symmetric_difference return self._shallow_copy_with_infer(the_diff, attribs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 387, in _shallow_copy_with_infer return self._shallow_copy(values, kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 396, in _shallow_copy return MultiIndex.from_tuples(values, **kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 883, in from_tuples raise TypeError('Cannot infer number of levels from empty list') TypeError: Cannot infer number of levels from empty list ``` #### Expected Output ``` MultiIndex(levels=[[], []], labels = [[], []]) ``` #### output of `pd.show_versions()` ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.11.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 61 Stepping 4, GenuineIntel byteorder: little LC_ALL: None LANG: DA pandas: 0.18.1 nose: None pip: 8.1.1 setuptools: 20.7.0 Cython: 0.23 numpy: 1.10.4 scipy: 0.17.0 statsmodels: 0.6.1 xarray: None IPython: 4.2.0 sphinx: None patsy: 0.4.1 dateutil: 2.5.2 pytz: 2016.3 blosc: None bottleneck: None tables: 3.2.2 numexpr: 2.5.2 matplotlib: 1.5.1 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None httplib2: None apiclient: None sqlalchemy: 1.0.12 pymysql: None psycopg2: None jinja2: 2.8 boto: None pandas_datareader: None ```	yeah I agree, on empties this should behave the same as `.difference` / `.union` cc @TomAugspurger Agreed.	2017-05-24T20:40:41Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1674, in __xor__ return self.symmetric_difference(other) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1911, in symmetric_difference return self._shallow_copy_with_infer(the_diff, attribs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 387, in _shallow_copy_with_infer return self._shallow_copy(values, kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 396, in _shallow_copy return MultiIndex.from_tuples(values, **kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 883, in from_tuples raise TypeError('Cannot infer number of levels from empty list') TypeError: Cannot infer number of levels from empty list	11,222
pandas-dev/pandas	pandas-dev__pandas-16526	ef487d9e474e8052c0f7c6260de5802a950defad		Various py3k test failures in tests.io.test_html with US-ASCII preferred encoding #### Code Sample, a copy-pastable example if possible ```python >>> import locale >>> locale.getpreferredencoding() 'US-ASCII' >>> open('/usr/local/lib/python3.4/site-packages/pandas/tests/io/data/spam.html').read() Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python3.4/encodings/ascii.py", line 26, in decode return codecs.ascii_decode(input, self.errors)[0] UnicodeDecodeError: 'ascii' codec can't decode byte 0xc2 in position 21552: ordinal not in range(128) >>> ``` #### Problem description Three tests, `test_string_io`, `test_string`, and `test_file_like`, all open `spam.html` without specifying the encoding, and then attempt to read it. This causes the tests to terminate prematurely with an error. #### Expected Output All three tests should pass since the code under test is not responsible for determining the file encoding. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.4.6.final.0 python-bits: 64 OS: FreeBSD OS-release: 10.3-STABLE machine: amd64 processor: amd64 byteorder: little LC_ALL: None LANG: None LOCALE: None.None pandas: 0.20.1 pytest: 3.1.0 pip: None setuptools: 32.1.0 Cython: None numpy: 1.11.2 scipy: 0.19.0 xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.0.0 tables: 3.4.2 numexpr: 2.6.2 feather: None matplotlib: None openpyxl: 2.4.7 xlrd: 1.0.0 xlwt: None xlsxwriter: 0.9.6 lxml: 3.6.0 bs4: 4.5.1 html5lib: 0.9999999 sqlalchemy: 1.1.10 pymysql: 0.7.11.None psycopg2: 2.7.1 (dt dec pq3 ext lo64) jinja2: 2.9.5 s3fs: None pandas_gbq: None pandas_datareader: None </details>		2017-05-28T18:59:57Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python3.4/encodings/ascii.py", line 26, in decode return codecs.ascii_decode(input, self.errors)[0] UnicodeDecodeError: 'ascii' codec can't decode byte 0xc2 in position 21552: ordinal not in range(128)	11,231
pandas-dev/pandas	pandas-dev__pandas-16543	03d44f3dd0ffd55d7538b67466cf4d3899ceac27	diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -44,8 +44,7 @@ Bug Fixes - Bug in ``DataFrame.update()`` with ``overwrite=False`` and ``NaN values`` (:issue:`15593`) - Passing an invalid engine to :func:`read_csv` now raises an informative ``ValueError`` rather than ``UnboundLocalError``. (:issue:`16511`) - - +- Bug in :func:`unique` on an array of tuples (:issue:`16519`) - Fixed a compatibility issue with IPython 6.0's tab completion showing deprecation warnings on Categoricals (:issue:`16409`) diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -163,7 +163,7 @@ def _ensure_arraylike(values): ABCIndexClass, ABCSeries)): inferred = lib.infer_dtype(values) if inferred in ['mixed', 'string', 'unicode']: - values = np.asarray(values, dtype=object) + values = lib.list_to_object_array(values) else: values = np.asarray(values) return values @@ -328,6 +328,11 @@ def unique(values): [b, a, c] Categories (3, object): [a < b < c] + An array of tuples + + >>> pd.unique([('a', 'b'), ('b', 'a'), ('a', 'c'), ('b', 'a')]) + array([('a', 'b'), ('b', 'a'), ('a', 'c')], dtype=object) + See Also -------- pandas.Index.unique	Regression from 0.19.2 to 0.20.1 in pandas.unique() when applied to list of tuples #### Code Sample, a copy-pastable example if possible ```python import pandas as pd input = [(0, 0), (0, 1), (1, 0), (1, 1), (0, 0), (0, 1), (1, 0), (1, 1)] print pd.unique(input) ``` #### Problem description The code exits unexpectedly ``` Traceback (most recent call last): File "pandas_bug.py", line 6, in <module> pd.unique(input) File "/Users/johannes/.virtualenvs/pandas/lib/python2.7/site-packages/pandas/core/algorithms.py", line 351, in unique uniques = table.unique(values) File "pandas/_libs/hashtable_class_helper.pxi", line 1271, in pandas._libs.hashtable.PyObjectHashTable.unique (pandas/_libs/hashtable.c:21384) ValueError: Buffer has wrong number of dimensions (expected 1, got 2) ``` #### Expected Output The code works on pandas version 0.19.2 and produces the expected output ``` [(0, 0) (0, 1) (1, 0) (1, 1)] ``` Moreover this problem is not limited to MacOSX, but was also encounter on Ubuntu CI server. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Darwin OS-release: 15.6.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: None.None pandas: 0.20.1 pytest: None pip: 9.0.1 setuptools: 35.0.2 Cython: None numpy: 1.12.1 scipy: None xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None None </details>	this is related to #16394 and needs the same fix, along with some tests; ensuring that nothing else breaks. ``` diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py index 77d79c9..9cfaf04 100644 --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -163,7 +163,7 @@ def _ensure_arraylike(values): ABCIndexClass, ABCSeries)): inferred = lib.infer_dtype(values) if inferred in ['mixed', 'string', 'unicode']: - values = np.asarray(values, dtype=object) + values = lib.list_to_object_array(values) else: values = np.asarray(values) return values ```	2017-05-30T22:05:37Z	[]	[]	Traceback (most recent call last): File "pandas_bug.py", line 6, in <module> pd.unique(input) File "/Users/johannes/.virtualenvs/pandas/lib/python2.7/site-packages/pandas/core/algorithms.py", line 351, in unique uniques = table.unique(values) File "pandas/_libs/hashtable_class_helper.pxi", line 1271, in pandas._libs.hashtable.PyObjectHashTable.unique (pandas/_libs/hashtable.c:21384) ValueError: Buffer has wrong number of dimensions (expected 1, got 2)	11,236
pandas-dev/pandas	pandas-dev__pandas-16701	b7e7fd3f17d4d2a2f87b9d169cf87143f04e5d33	diff --git a/doc/source/whatsnew/v0.20.3.txt b/doc/source/whatsnew/v0.20.3.txt --- a/doc/source/whatsnew/v0.20.3.txt +++ b/doc/source/whatsnew/v0.20.3.txt @@ -59,7 +59,7 @@ I/O Plotting ^^^^^^^^ - +- Fix regression in series plotting that prevented RGB and RGBA tuples from being used as color arguments (:issue:`16233`) diff --git a/pandas/plotting/_core.py b/pandas/plotting/_core.py --- a/pandas/plotting/_core.py +++ b/pandas/plotting/_core.py @@ -187,6 +187,11 @@ def _validate_color_args(self): # support series.plot(color='green') self.kwds['color'] = [self.kwds['color']] + if ('color' in self.kwds and isinstance(self.kwds['color'], tuple) and + self.nseries == 1 and len(self.kwds['color']) in (3, 4)): + # support RGB and RGBA tuples in series plot + self.kwds['color'] = [self.kwds['color']] + if ('color' in self.kwds or 'colors' in self.kwds) and \ self.colormap is not None: warnings.warn("'color' and 'colormap' cannot be used "	DataFrame plot method no long takes RGB tuple as color arg #### Problem description The `plot` method on `DataFrame` objects takes a `color` argument that in versions prior to 0.20.2 took an RGB tuple as an accepted value. The 0.20.2 release throws an exception when specifying an RGB tuple for the `color` arg. #### Code Sample ```python # import matplotlib.pyplot as plt df = pandas.DataFrame([[1, 2], [3, 4]], columns=['a', 'b']) df.plot(x='b', y='a', color=(1, 0, 0)) plt.show() ``` #### Expected Output The expected output is a red line plot. The following exception is thrown: ```python Exception in Tkinter callback Traceback (most recent call last): File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/lib-tk/Tkinter.py", line 1536, in __call__ return self.func(args) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 280, in resize self.show() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 351, in draw FigureCanvasAgg.draw(self) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 464, in draw self.figure.draw(self.renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, args, *kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/figure.py", line 1144, in draw renderer, self, dsu, self.suppressComposite) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, args, *kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/axes/_base.py", line 2426, in draw mimage._draw_list_compositing_images(renderer, self, dsu) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 803, in draw ln_color_rgba = self._get_rgba_ln_color() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 1344, in _get_rgba_ln_color return mcolors.to_rgba(self._color, self._alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 143, in to_rgba rgba = _to_rgba_no_colorcycle(c, alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 194, in _to_rgba_no_colorcycle raise ValueError("Invalid RGBA argument: {!r}".format(orig_c)) ValueError: Invalid RGBA argument: 1 ``` This plot is correctly displayed with an identical environment that has pandas 0.19.2 installed so is API breaking. Apologies if this is fixed already in master. #### Output of ``pd.show_versions()`` <details> NSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Darwin OS-release: 15.2.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.20.2 pytest: None pip: 9.0.1 setuptools: 36.0.1 Cython: None numpy: 1.13.0 scipy: None xarray: None IPython: 5.4.1 sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: 2.0.2 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None </details>	Follow up: Enclosing the tuple in a list works: ``` df.plot(x='b', y='a', color=[(1, 0, 0)]) ``` This looks related to #16233 which fixes #3486 but breaks what I perceive to be a fairly standard way of assigning line color (e.g. it is the first method listed for specifying color on https://matplotlib.org/users/colors.html). @kjford Thanks for the report! That is indeed a regression, and should be fixed. Want to do a PR to fix?	2017-06-15T06:33:34Z	[]	[]	Traceback (most recent call last): File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/lib-tk/Tkinter.py", line 1536, in __call__ return self.func(args) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 280, in resize self.show() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 351, in draw FigureCanvasAgg.draw(self) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 464, in draw self.figure.draw(self.renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, args, *kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/figure.py", line 1144, in draw renderer, self, dsu, self.suppressComposite) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, args, *kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/axes/_base.py", line 2426, in draw mimage._draw_list_compositing_images(renderer, self, dsu) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 803, in draw ln_color_rgba = self._get_rgba_ln_color() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 1344, in _get_rgba_ln_color return mcolors.to_rgba(self._color, self._alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 143, in to_rgba rgba = _to_rgba_no_colorcycle(c, alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 194, in _to_rgba_no_colorcycle raise ValueError("Invalid RGBA argument: {!r}".format(orig_c)) ValueError: Invalid RGBA argument: 1	11,258
pandas-dev/pandas	pandas-dev__pandas-16744	8a98f5ed541c87a9bf101c9331bd6cfa8f007cc9		Different behaviour on two different environments. TypeError: data type "datetime" not understood #### I have an aggregation logic ```python In [19]: datewise_prices Out[19]: [{'arrivalDate': 1490227200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490400000000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490832000000, 'maxPrice': 3500, 'minPrice': 3500, 'modalPrice': 3500}, {'arrivalDate': 1490918400000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1491091200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}] In [20]: weekly_dataframe = pandas.DataFrame(datewise_prices) ...: weekly_dataframe.drop('minPrice', axis=1, inplace=True) ...: weekly_dataframe.drop('maxPrice', axis=1, inplace=True) ...: weekly_dataframe['arrivalDate'] = pandas.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') ...: weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() ...: weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arriv ...: alDate'].dt.strftime('%d %b') ``` #### It results in an error on my Staging environment but works fine on my local. The staging environment gives an error in `(weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b')` ``` Traceback (most recent call last): File "/root/myapp/myapp/handlers/aggregation_handler.py", line 290, in get_daily_weekly_and_monthly_aggregates weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 696, in wrapper converted = _Op.get_op(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 330, in get_op return _TimeOp(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 343, in __init__ lvalues = self._convert_to_array(left, name=name) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 454, in _convert_to_array if (inferred_type in ('datetime64', 'datetime', 'date', 'time') or TypeError: data type "datetime" not understood ``` but it works fine on my local. #### Output of ``pd.show_versions()`` on local <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.6.final.0 python-bits: 64 OS: Linux OS-release: 4.4.0-78-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_IN LOCALE: None.None pandas: 0.19.2 nose: None pip: 9.0.1 setuptools: 36.0.1 Cython: None numpy: 1.12.1 scipy: None statsmodels: None xarray: None IPython: 5.4.1 sphinx: None patsy: None dateutil: 2.5.3 pytz: 2016.7 blosc: None bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: 0.9.6 lxml: None bs4: None html5lib: 0.999 httplib2: 0.8 apiclient: None sqlalchemy: None pymysql: None psycopg2: None jinja2: 2.8 boto: 2.40.0 pandas_datareader: None </details> #### Output of ``pd.show_versions()`` on Staging <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.6.final.0 python-bits: 64 OS: Linux OS-release: 3.13.0-57-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: None.None pandas: 0.19.2 nose: None pip: 1.5.4 setuptools: 3.3 Cython: None numpy: 1.13.0 scipy: None statsmodels: None xarray: None IPython: 4.0.0 sphinx: None patsy: None dateutil: 2.5.3 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: 0.9.6 lxml: None bs4: None html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.8 pymysql: None psycopg2: None jinja2: 2.9.6 boto: 2.40.0 pandas_datareader: None </details> Please help me out.	Can you make a copy-pastable example, as requested? ``` import pandas as pd datewise_prices = [ {'arrivalDate': 1490227200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490400000000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490832000000, 'maxPrice': 3500, 'minPrice': 3500, 'modalPrice': 3500}, {'arrivalDate': 1490918400000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1491091200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300} ] weekly_dataframe = pd.DataFrame(datewise_prices) weekly_dataframe.drop('minPrice', axis=1, inplace=True) weekly_dataframe.drop('maxPrice', axis=1, inplace=True) weekly_dataframe['arrivalDate'] = pd.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pd.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') ``` Transferred from @hussaintamboli 's input. It will help if you provide more information, for example, the full trace when it error. @BranYang and @TomAugspurger, I have added the stacktrace. Just figured it out myself. It's because of numpy==1.13.0. For numpy==1.12.1, it's working fine. this looks fine. what exactly is the issue? On staging environment with the dependencies mentioned in `Output of pd.show_versions() on Staging`, I get `TypeError: data type "datetime" not understood`. But on local environment with dependencies mentioned in `Output of pd.show_versions() on local`, I get desired output. your example does not repro. I would guess that you have strings (and not datetimes) in one of your fields Reproduction steps. This works ``` $ pip freeze \| grep numpy numpy==1.12.1 import pandas as pd datewise_prices = [ {'arrivalDate': 1490227200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490400000000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490832000000, 'maxPrice': 3500, 'minPrice': 3500, 'modalPrice': 3500}, {'arrivalDate': 1490918400000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1491091200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300} ] weekly_dataframe = pd.DataFrame(datewise_prices) weekly_dataframe.drop('minPrice', axis=1, inplace=True) weekly_dataframe.drop('maxPrice', axis=1, inplace=True) weekly_dataframe['arrivalDate'] = pd.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pd.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') weekly_dataframe.drop('arrivalDate', axis=1, inplace=True) print weekly_dataframe.to_dict(orient='records') [{'label': u'20 Mar to 26 Mar', 'modalPrice': 3300.0}, {'label': u'27 Mar to 02 Apr', 'modalPrice': 3366.6666666666665}] ``` On the staging environment where it gives error ``` pip install numpy==1.13.0 weekly_dataframe = pd.DataFrame(datewise_prices) weekly_dataframe.drop('minPrice', axis=1, inplace=True) weekly_dataframe.drop('maxPrice', axis=1, inplace=True) weekly_dataframe['arrivalDate'] = pd.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pd.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') TypeError: data type "datetime" not understood ``` this works on master for both numpies so not sure what your issue is I can reproduce it on 0.19.2 / 1.13.0. Small reproducible example: ``` pd.Series(pd.date_range("2012-01-01", periods=3)) - pd.offsets.DateOffset(days=6) ``` and the error actually comes from ``` np.dtype('M8[ns]') in ('datetime', 'datetime64') ``` But on master / 0.20.2, this does not raise anymore, so I suppose we have already fixed this (this line of code is also removed). Nonetheless, always welcome to do a PR to add a test for this specific case (the small reproducible example I showed above), to make sure it keeps on working. yeah we had some refactorings of this in 0.20.0. ok test PR it is. @hussaintamboli Would you be interested to add such a test? Sure. I'll do it.	2017-06-21T11:24:41Z	[]	[]	Traceback (most recent call last): File "/root/myapp/myapp/handlers/aggregation_handler.py", line 290, in get_daily_weekly_and_monthly_aggregates weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 696, in wrapper converted = _Op.get_op(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 330, in get_op return _TimeOp(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 343, in __init__ lvalues = self._convert_to_array(left, name=name) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 454, in _convert_to_array if (inferred_type in ('datetime64', 'datetime', 'date', 'time') or TypeError: data type "datetime" not understood	11,263
pandas-dev/pandas	pandas-dev__pandas-16926	a587d568d213c62307a72d98d6913239f55844e8	diff --git a/doc/source/whatsnew/v0.21.0.txt b/doc/source/whatsnew/v0.21.0.txt --- a/doc/source/whatsnew/v0.21.0.txt +++ b/doc/source/whatsnew/v0.21.0.txt @@ -162,6 +162,7 @@ I/O - Bug in :func:`read_csv` in which non integer values for the header argument generated an unhelpful / unrelated error message (:issue:`16338`) +- Bug in :func:`read_stata` where value labels could not be read when using an iterator (:issue:`16923`) Plotting ^^^^^^^^ diff --git a/pandas/io/stata.py b/pandas/io/stata.py --- a/pandas/io/stata.py +++ b/pandas/io/stata.py @@ -997,6 +997,7 @@ def __init__(self, path_or_buf, convert_dates=True, self.path_or_buf = BytesIO(contents) self._read_header() + self._setup_dtype() def __enter__(self): """ enter context manager """ @@ -1299,6 +1300,23 @@ def _read_old_header(self, first_char): # necessary data to continue parsing self.data_location = self.path_or_buf.tell() + def _setup_dtype(self): + """Map between numpy and state dtypes""" + if self._dtype is not None: + return self._dtype + + dtype = [] # Convert struct data types to numpy data type + for i, typ in enumerate(self.typlist): + if typ in self.NUMPY_TYPE_MAP: + dtype.append(('s' + str(i), self.byteorder + + self.NUMPY_TYPE_MAP[typ])) + else: + dtype.append(('s' + str(i), 'S' + str(typ))) + dtype = np.dtype(dtype) + self._dtype = dtype + + return self._dtype + def _calcsize(self, fmt): return (type(fmt) is int and fmt or struct.calcsize(self.byteorder + fmt)) @@ -1472,22 +1490,10 @@ def read(self, nrows=None, convert_dates=None, if nrows is None: nrows = self.nobs - if (self.format_version >= 117) and (self._dtype is None): + if (self.format_version >= 117) and (not self._value_labels_read): self._can_read_value_labels = True self._read_strls() - # Setup the dtype. - if self._dtype is None: - dtype = [] # Convert struct data types to numpy data type - for i, typ in enumerate(self.typlist): - if typ in self.NUMPY_TYPE_MAP: - dtype.append(('s' + str(i), self.byteorder + - self.NUMPY_TYPE_MAP[typ])) - else: - dtype.append(('s' + str(i), 'S' + str(typ))) - dtype = np.dtype(dtype) - self._dtype = dtype - # Read data dtype = self._dtype max_read_len = (self.nobs - self._lines_read) * dtype.itemsize @@ -1958,7 +1964,6 @@ def _prepare_categoricals(self, data): return data get_base_missing_value = StataMissingValue.get_base_missing_value - index = data.index data_formatted = [] for col, col_is_cat in zip(data, is_cat): if col_is_cat: @@ -1981,8 +1986,7 @@ def _prepare_categoricals(self, data): # Replace missing values with Stata missing value for type values[values == -1] = get_base_missing_value(dtype) - data_formatted.append((col, values, index)) - + data_formatted.append((col, values)) else: data_formatted.append((col, data[col])) return DataFrame.from_items(data_formatted)	Unable to read Stata value_labels from .dta-file created by pandas (to_stata()) #### Code Sample, a copy-pastable example if possible ```python import pandas as pd d = {'A':['B','E','C','A','E']} df = pd.DataFrame(data=d) df['A'] = df['A'].astype('category') # Setting as categorical, similar to value_label in Stata df.to_stata('test.dta') # Writing dataframe to Stata-file dfs_fromstata = pd.read_stata('test.dta', iterator=True) # Creating StataReader-object print(dfs_fromstata.value_labels()) # Printing value_labels ``` ```python Traceback (most recent call last): File "<stdin>", line 1, in <module> print(dfs_fromstata.value_labels()) # Printing value_labels File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1725, in value_labels self._read_value_labels() File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1329, in _read_value_labels offset = self.nobs * self._dtype.itemsize AttributeError: 'NoneType' object has no attribute 'itemsize' ``` #### Problem description It seems as if read_stata() is not able to read value_labels properly from a .dta-file created by to_stata(). If the file is created by Stata itself, the value_labels are read correctly. Also, if the .dta-file created by to_stata() is opened and saved by Stata, the value_labels are read correctly. #### Expected Output {'A': {0: 'A', 1: 'B', 2: 'C', 3: 'E'}} #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.6.0.final.0 python-bits: 64 OS: Windows OS-release: 10 machine: AMD64 processor: Intel64 Family 6 Model 78 Stepping 3, GenuineIntel byteorder: little LC_ALL: None LANG: None LOCALE: None.None pandas: 0.20.2 pytest: 3.0.5 pip: 9.0.1 setuptools: 27.2.0 Cython: 0.25.2 numpy: 1.12.1 scipy: 0.19.1 xarray: None IPython: 5.1.0 sphinx: 1.5.1 patsy: 0.4.1 dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.2.1 tables: 3.2.2 numexpr: 2.6.2 feather: None matplotlib: 2.0.2 openpyxl: 2.4.1 xlrd: 1.0.0 xlwt: 1.2.0 xlsxwriter: 0.9.6 lxml: 3.7.2 bs4: 4.5.3 html5lib: None sqlalchemy: 1.1.5 pymysql: None psycopg2: None jinja2: 2.9.4 s3fs: None pandas_gbq: None pandas_datareader: None </details>	cc @bashtage I don't think this is supported. It works fine. You need to read once in order to initialize some values. This appears to be an omission in the iterator implementation. Should probably raise with a usable error in the scenario. ``` import pandas as pd d = {'A':['B','E','C','A','E']} df = pd.DataFrame(data=d) df['A'] = df['A'].astype('category') # Setting as categorical, similar to value_label in Stata df.to_stata('test.dta') # Writing dataframe to Stata-file dfs_fromstata = pd.read_stata('test.dta', iterator=True) # Creating StataReader-object dfs_fromstata.read() print(dfs_fromstata.value_labels()) ``` pull-requests welcome! Should pull this block https://github.com/pandas-dev/pandas/blob/master/pandas/io/stata.py#L1480 out to a stand alone function and then call it once the header has been read.	2017-07-14T17:22:36Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> print(dfs_fromstata.value_labels()) # Printing value_labels File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1725, in value_labels self._read_value_labels() File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1329, in _read_value_labels offset = self.nobs * self._dtype.itemsize AttributeError: 'NoneType' object has no attribute 'itemsize'	11,285
pandas-dev/pandas	pandas-dev__pandas-16930	4efe6560e07f28de6a1834fa90e31cef31b0fb18	diff --git a/doc/source/whatsnew/v0.21.0.txt b/doc/source/whatsnew/v0.21.0.txt --- a/doc/source/whatsnew/v0.21.0.txt +++ b/doc/source/whatsnew/v0.21.0.txt @@ -164,6 +164,8 @@ I/O - Bug in :func:`read_stata` where value labels could not be read when using an iterator (:issue:`16923`) +- Bug in :func:`read_html` where import check fails when run in multiple threads (:issue:`16928`) + Plotting ^^^^^^^^ - Bug in plotting methods using ``secondary_y`` and ``fontsize`` not setting secondary axis font size (:issue:`12565`) diff --git a/pandas/io/html.py b/pandas/io/html.py --- a/pandas/io/html.py +++ b/pandas/io/html.py @@ -37,8 +37,6 @@ def _importers(): if _IMPORTS: return - _IMPORTS = True - global _HAS_BS4, _HAS_LXML, _HAS_HTML5LIB try: @@ -59,6 +57,8 @@ def _importers(): except ImportError: pass + _IMPORTS = True + ############# # READ HTML #	read_html() Thread Safety #### Code Sample ```python #!/usr/bin/python3 import pandas import threading def fetch_file(): url = "https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.html" pandas.read_html(url) thread1 = threading.Thread(target = fetch_file) thread2 = threading.Thread(target = fetch_file) thread1.start() thread2.start() ``` ### Output ``` Exception in thread Thread-2: Traceback (most recent call last): File "/usr/lib/python3.6/threading.py", line 916, in _bootstrap_inner self.run() File "/usr/lib/python3.6/threading.py", line 864, in run self._target(self._args, *self._kwargs) File "./pandas_bug.py", line 7, in fetch_file pandas.read_html(url) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 904, in read_html keep_default_na=keep_default_na) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 731, in _parse parser = _parser_dispatch(flav) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 691, in _parser_dispatch raise ImportError("lxml not found, please install it") ImportError: lxml not found, please install it ``` #### Problem description read_html() doesn't appear to be multi-threading safe. This specific issue seems to be caused by setting `_IMPORTS` in html.py to True too early resulting in the second thread entering `_parser_dispatch` and throwing an exception while the first thread hasn't finished the check. I have written a potential fix and will open a PR shortly. #### Expected Output No exception should be thrown since lxml is installed and the program works fine without multi-threading. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------------ commit: None python: 3.6.1.final.0 python-bits: 64 OS: Linux OS-release: 4.11.3-1-ARCH machine: x86_64 processor: byteorder: little LC_ALL: None LANG: en_GB.UTF-8 LOCALE: en_GB.UTF-8 pandas: 0.20.1 pytest: None pip: 9.0.1 setuptools: 36.0.1 Cython: None numpy: 1.12.1 scipy: 0.19.0 xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: 2.0.2 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: 4.6.0 html5lib: 0.999999999 sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None </details>		2017-07-14T19:25:11Z	[]	[]	Traceback (most recent call last): File "/usr/lib/python3.6/threading.py", line 916, in _bootstrap_inner self.run() File "/usr/lib/python3.6/threading.py", line 864, in run self._target(self._args, *self._kwargs) File "./pandas_bug.py", line 7, in fetch_file pandas.read_html(url) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 904, in read_html keep_default_na=keep_default_na) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 731, in _parse parser = _parser_dispatch(flav) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 691, in _parser_dispatch raise ImportError("lxml not found, please install it") ImportError: lxml not found, please install it	11,286
pandas-dev/pandas	pandas-dev__pandas-17169	9b07ef4a5b656a1532512c270533053ee338e30d	diff --git a/pandas/core/internals.py b/pandas/core/internals.py --- a/pandas/core/internals.py +++ b/pandas/core/internals.py @@ -323,10 +323,6 @@ def reindex_axis(self, indexer, method=None, axis=1, fill_value=None, fill_value=fill_value, mask_info=mask_info) return self.make_block(new_values, fastpath=True) - def get(self, item): - loc = self.items.get_loc(item) - return self.values[loc] - def iget(self, i): return self.values[i] @@ -1658,13 +1654,6 @@ def set(self, locs, values, check=False): assert locs.tolist() == [0] self.values = values - def get(self, item): - if self.ndim == 1: - loc = self.items.get_loc(item) - return self.values[loc] - else: - return self.values - def putmask(self, mask, new, align=True, inplace=False, axis=0, transpose=False, mgr=None): """ @@ -4722,8 +4711,6 @@ def _concat_indexes(indexes): def _block2d_to_blocknd(values, placement, shape, labels, ref_items): """ pivot to the labels shape """ - from pandas.core.internals import make_block - panel_shape = (len(placement),) + shape # TODO: lexsort depth needs to be 2!!	AttributeError Block.items https://github.com/pandas-dev/pandas/blob/master/pandas/core/internals.py#L326 `core.internals.Block` references `self.items`. AFAICT `items` is an attribute of `BlockManager`, does not exist in `Block`. ``` ser = pd.Series(range(5)) mgr = ser._data block = mgr.blocks[0] >>> block.get(1) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python2.7/site-packages/pandas/core/internals.py", line 329, in get loc = self.items.get_loc(item) AttributeError: 'IntBlock' object has no attribute 'items' ```	I think this is dead code. We always use positional indexing on a block (e.g. ``.iget``). not sure why the linter doesn't find this. you can do a PR to remove. Code coverage also indicates this is never used (at least in our tests)	2017-08-04T07:13:48Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python2.7/site-packages/pandas/core/internals.py", line 329, in get loc = self.items.get_loc(item) AttributeError: 'IntBlock' object has no attribute 'items'	11,323
pandas-dev/pandas	pandas-dev__pandas-17194	3c833db29b6f5977c78d1ade791a09a5b29cedb8	diff --git a/doc/source/whatsnew/v0.21.0.txt b/doc/source/whatsnew/v0.21.0.txt --- a/doc/source/whatsnew/v0.21.0.txt +++ b/doc/source/whatsnew/v0.21.0.txt @@ -278,6 +278,7 @@ Indexing - Fixes bug where indexing with ``np.inf`` caused an ``OverflowError`` to be raised (:issue:`16957`) - Bug in reindexing on an empty ``CategoricalIndex`` (:issue:`16770`) - Fixes ``DataFrame.loc`` for setting with alignment and tz-aware ``DatetimeIndex`` (:issue:`16889`) +- Avoids ``IndexError`` when passing an Index or Series to ``.iloc`` with older numpy (:issue:`17193`) I/O ^^^ diff --git a/pandas/core/internals.py b/pandas/core/internals.py --- a/pandas/core/internals.py +++ b/pandas/core/internals.py @@ -861,6 +861,9 @@ def _is_empty_indexer(indexer): # set else: + if _np_version_under1p9: + # Work around GH 6168 to support old numpy + indexer = getattr(indexer, 'values', indexer) values[indexer] = value # coerce and try to infer the dtypes of the result	passing Index or Series to iloc fails with old numpy #### Code Sample, a copy-pastable example if possible ```python >>> s = pd.Series([1,2]) >>> s.iloc[pd.Series([0])] = 2 Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/indexing.py", line 198, in __setitem__ self._setitem_with_indexer(indexer, value) File "pandas/core/indexing.py", line 619, in _setitem_with_indexer value=value) File "pandas/core/internals.py", line 3313, in setitem return self.apply('setitem', kwargs) File "pandas/core/internals.py", line 3201, in apply applied = getattr(b, f)(kwargs) File "pandas/core/internals.py", line 864, in setitem values[indexer] = value IndexError: unsupported iterator index >>> np.version.version '1.7.0' ``` #### Problem description This is a consequence of #6168 , which is fixed in more recent numpy versions (but 1.7.0 is still a supported version). Not 100% sure this is worth fixing (i.e. how interested we are in keeping compatibility with numpy 1.8.0), but the fix is trivial, so I'll just push a PR and let you judge. #### Expected Output Just no error. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.3.final.0 python-bits: 64 OS: Linux OS-release: 4.9.0-3-amd64 machine: x86_64 processor: byteorder: little LC_ALL: None LANG: it_IT.UTF-8 LOCALE: None.None pandas: 0+unknown pytest: 3.2.0 pip: None setuptools: 36.2.7 Cython: None numpy: 1.7.0 scipy: None pyarrow: None xarray: None IPython: None sphinx: None patsy: None dateutil: 1.5 pytz: 2012c blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None fastparquet: None pandas_gbq: None pandas_datareader: None >>> </details>		2017-08-07T22:07:19Z	[]	[]	Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/indexing.py", line 198, in __setitem__ self._setitem_with_indexer(indexer, value) File "pandas/core/indexing.py", line 619, in _setitem_with_indexer value=value) File "pandas/core/internals.py", line 3313, in setitem return self.apply('setitem', kwargs) File "pandas/core/internals.py", line 3201, in apply applied = getattr(b, f)(kwargs) File "pandas/core/internals.py", line 864, in setitem values[indexer] = value IndexError: unsupported iterator index	11,325
pandas-dev/pandas	pandas-dev__pandas-17201	674fb96b33c07c680844f674fcdf0767b6e3c2f9	diff --git a/doc/source/whatsnew/v0.21.1.txt b/doc/source/whatsnew/v0.21.1.txt --- a/doc/source/whatsnew/v0.21.1.txt +++ b/doc/source/whatsnew/v0.21.1.txt @@ -88,7 +88,7 @@ I/O - :func:`read_parquet` now allows to specify kwargs which are passed to the respective engine (:issue:`18216`) - Bug in parsing integer datetime-like columns with specified format in ``read_sql`` (:issue:`17855`). - Bug in :meth:`DataFrame.to_msgpack` when serializing data of the numpy.bool_ datatype (:issue:`18390`) - +- Bug in :func:`read_json` not decoding when reading line deliminted JSON from S3 (:issue:`17200`) Plotting ^^^^^^^^ diff --git a/pandas/io/json/json.py b/pandas/io/json/json.py --- a/pandas/io/json/json.py +++ b/pandas/io/json/json.py @@ -5,7 +5,7 @@ import pandas._libs.json as json from pandas._libs.tslib import iNaT -from pandas.compat import StringIO, long, u +from pandas.compat import StringIO, long, u, to_str from pandas import compat, isna from pandas import Series, DataFrame, to_datetime, MultiIndex from pandas.io.common import (get_filepath_or_buffer, _get_handle, @@ -458,8 +458,10 @@ def read(self): if self.lines and self.chunksize: obj = concat(self) elif self.lines: + + data = to_str(self.data) obj = self._get_object_parser( - self._combine_lines(self.data.split('\n')) + self._combine_lines(data.split('\n')) ) else: obj = self._get_object_parser(self.data) @@ -612,7 +614,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: dtype = np.dtype(dtype) return data.astype(dtype), True - except: + except (TypeError, ValueError): return data, False if convert_dates: @@ -628,7 +630,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: data = data.astype('float64') result = True - except: + except (TypeError, ValueError): pass if data.dtype.kind == 'f': @@ -639,7 +641,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: data = data.astype('float64') result = True - except: + except (TypeError, ValueError): pass # do't coerce 0-len data @@ -651,7 +653,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, if (new_data == data).all(): data = new_data result = True - except: + except (TypeError, ValueError): pass # coerce ints to 64 @@ -661,7 +663,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: data = data.astype('int64') result = True - except: + except (TypeError, ValueError): pass return data, result @@ -680,7 +682,7 @@ def _try_convert_to_date(self, data): if new_data.dtype == 'object': try: new_data = data.astype('int64') - except: + except (TypeError, ValueError): pass # ignore numbers that are out of range @@ -697,7 +699,7 @@ def _try_convert_to_date(self, data): unit=date_unit) except ValueError: continue - except: + except Exception: break return new_data, True return data, False	read_json(lines=True) broken for s3 urls in Python 3 (v0.20.3) #### Code Sample, a copy-pastable example if possible Using Python ```python import pandas as pd inputdf = pd.read_json(path_or_buf="s3://path/to/python-lines/file.json", lines=True) ``` The file is similar to: ``` {"url": "blah", "other": "blah"} {"url": "blah", "other": "blah"} {"url": "blah", "other": "blah"} ``` #### Problem description When attempting to read a python lines file into a DataFrame using the s3 protocol, the above code will error with: ``` 2017-08-08 11:06:14,225 - image_rank_csv - ERROR - initial_value must be str or None, not bytes Traceback (most recent call last): File "image_rank_csv.py", line 62, in run inputdf = pd.read_json(path_or_buf="s3://path/to/python-lines/file.json", lines=True) File "...env/lib/python3.6/site-packages/pandas/io/json/json.py", line 347, in read_json lines = list(StringIO(json.strip())) TypeError: initial_value must be str or None, not bytes ``` This works fine if the file is local, e.g.: ```python import pandas as pd inputdf = pd.read_json(path_or_buf="/local/path/to/python-lines/file.json", lines=True) ``` #### Expected Output Expect to successfully read the file and error above not to occur. My current thinking is that when we get the file handle: https://github.com/pandas-dev/pandas/blob/v0.20.3/pandas/io/json/json.py#L333 , you delegate to `s3fs`, which documents that [it only operates in Binary mode](http://s3fs.readthedocs.io/en/latest/#limitations). Therefore when you `read()`: https://github.com/pandas-dev/pandas/blob/v0.20.3/pandas/io/json/json.py#L335, Therefore passing to `StringIO` will fail here: https://github.com/pandas-dev/pandas/blob/v0.20.3/pandas/io/json/json.py#L347 . Maybe it needs a different handler for `BytesIO`? #### Output of ``pd.show_versions()`` <details> [paste the output of ``pd.show_versions()`` here below this line] ``` INSTALLED VERSIONS ------------------ commit: None python: 3.6.1.final.0 python-bits: 64 OS: Darwin OS-release: 16.6.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.20.3 pytest: None pip: 9.0.1 setuptools: 36.2.7 Cython: None numpy: 1.12.0 scipy: 0.19.1 xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: 2.6.2 (dt dec pq3 ext lo64) jinja2: None s3fs: 0.1.2 pandas_gbq: None pandas_datareader: None ``` </details>	Well we do have a `BytesIO` class in `pandas.compat`. If we can condition on the data returned to us, that would be reasonable I think. @gfyoung I'm not intimately familiar with the codebase, but I have a (possibly naive) fix that just attempts to decode the json based on whatever `encoding` is. It seems to fix. Would you like to review if do a PR? > I'm not intimately familiar with the codebase Don't worry, that's a pretty tall order :smile: Absolutely! Submit a PR, and we'll certainly review it. FYI, here's how we handle it on the CSV side: https://github.com/pandas-dev/pandas/blob/3c833db29b6f5977c78d1ade791a09a5b29cedb8/pandas/io/common.py#L401 that `f` would be an instance of `S3File`, which is one of the classes in `need_text_wrapping`. By the `TextIOWrapper` can go around buffer of bytes and it'll do the encoding. You might be able to reuse parts of that for `read_json`, or just do something similar.	2017-08-08T18:39:45Z	[]	[]	Traceback (most recent call last): File "image_rank_csv.py", line 62, in run inputdf = pd.read_json(path_or_buf="s3://path/to/python-lines/file.json", lines=True) File "...env/lib/python3.6/site-packages/pandas/io/json/json.py", line 347, in read_json lines = list(StringIO(json.strip())) TypeError: initial_value must be str or None, not bytes	11,326

numpy/numpy

numpy__numpy-24299

b4c50e4862dd23ffcb27fe688d5e79e0456b2715

diff --git a/numpy/f2py/crackfortran.py b/numpy/f2py/crackfortran.py --- a/numpy/f2py/crackfortran.py +++ b/numpy/f2py/crackfortran.py @@ -1742,6 +1742,23 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: del d1[k] + if 'len' in d1 and 'array' in d1: + if d1['len'] == '': + d1['len'] = d1['array'] + del d1['array'] + elif typespec == 'character': + if ('charselector' not in edecl) or (not edecl['charselector']): + edecl['charselector'] = {} + if 'len' in edecl['charselector']: + del edecl['charselector']['len'] + edecl['charselector']['*'] = d1['len'] + del d1['len'] + else: + d1['array'] = d1['array'] + ',' + d1['len'] + del d1['len'] + errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( + typespec, e, typespec, ename, d1['array'])) + if 'len' in d1: if typespec in ['complex', 'integer', 'logical', 'real']: if ('kindselector' not in edecl) or (not edecl['kindselector']): @@ -1763,16 +1780,6 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: edecl['='] = d1['init'] - if 'len' in d1 and 'array' in d1: - if d1['len'] == '': - d1['len'] = d1['array'] - del d1['array'] - else: - d1['array'] = d1['array'] + ',' + d1['len'] - del d1['len'] - errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( - typespec, e, typespec, ename, d1['array'])) - if 'array' in d1: dm = 'dimension(%s)' % d1['array'] if 'attrspec' not in edecl or (not edecl['attrspec']):

BUG: f2py cannot compile files it used to be able to compile ### Describe the issue: I'm attempting to use f2py to compile some fortran code. I'm able to do this using the following numpy versions: - 1.21.0 - 1.22.0 - 1.23.0 - 1.24.0 - 1.24.2 Starting on v1.24.3, the same compilation code no longer works using f2py. Strangely enough, if I precompile one of files (LAPACK.f), the compilation can work with v1.24.3. ### Reproduce the code example: ```python # I don't see how I can produce runnable code using f2py, as it requires the fortran source code # This code produces a .so file in my project's home directory on numpy <= 1.24.2 but doesn't work on numpy >= 1.24.3 from pathlib import Path from numpy import f2py project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.f', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) # If I precompile LAPACK.f using: # /usr/bin/gfortran -Wall -g -ffixed-form -fno-second-underscore -g -fno-second-underscore -fPIC -O3 -funroll-loops -c LAPACK.f # then it works using numpy = 1.24.3. Note the only difference in the code is the LAPACK.f is now LAPACK.o project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.o', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) ``` ### Error message: ```shell There is no error message. f2py simply stops right before where it prints this line: INFO: compiling Fortran sources ``` ### Runtime information: Line 1 output: 1.24.0 3.10.6 (main, May 29 2023, 11:10:38) [GCC 11.3.0] (note that I have no idea why it says 1.24.0 when Pycharm assures me I'm using 1.24.2) Line 2 output: Exception ignored on calling ctypes callback function: <function ThreadpoolController._find_libraries_with_dl_iterate_phdr.<locals>.match_library_callback at 0x7f592f9c00d0> Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error [{'simd_extensions': {'baseline': ['SSE', 'SSE2', 'SSE3'], 'found': ['SSSE3', 'SSE41', 'POPCNT', 'SSE42', 'AVX', 'F16C', 'FMA3', 'AVX2'], 'not_found': ['AVX512F', 'AVX512CD', 'AVX512_KNL', 'AVX512_KNM', 'AVX512_SKX', 'AVX512_CLX', 'AVX512_CNL', 'AVX512_ICL']}}] None ### Context for the issue: I believe that LAPACK.f is one of the more widely used fortran codes. If it cannot compile in conjunction with other code, that could potentially disrupt a good number of users.

Ping @HaoZeke, also to confirm, can you check if the issue persists on 1.25.0? @seberg I can confirm that the issue persists on 1.25.0. It's actually why I noticed it in the first place. My code that used to run failed and then I tracked it down to the version number described above, in hopes that someone might have a better idea what caused it. On `1.24.4` the error is: ```bash ❯ f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f running build running config_cc INFO: unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc INFO: unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src INFO: build_src INFO: building extension "disort" sources INFO: f2py options: [] INFO: f2py:> /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c creating /tmp/tmpuul0531u/src.linux-x86_64-3.9 Reading fortran codes... Reading file 'BDREF.f' (format:fix,strict) Reading file 'DISOBRDF.f' (format:fix,strict) rmbadname1: Replacing "float" with "float_bn". rmbadname1: Replacing "len" with "len_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "float" with "float_bn". Reading file 'ERRPACK.f' (format:fix,strict) Reading file 'LAPACK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #745 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #1363 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #2115 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "char" with "char_bn". rmbadname1: Replacing "int" with "int_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #3999 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". Line #4273 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #5124 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #5462 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #6657 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Reading file 'LINPAK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Reading file 'RDI1MACH.f' (format:fix,strict) Post-processing... Block: disort Block: bdref Block: brdf_hapke Block: brdf_rpv Block: brdf_rossli Block: oceabrdf2 Block: shadow_eta Block: disobrdf {} In: :disort:DISOBRDF.f:surfac2 vars2fortran: No typespec for argument "nazz". Block: surfac2 Block: qgausn2 Block: zeroit2 Block: errmsg Block: wrtbad Block: wrtdim Block: tstbad Block: dgemm In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dger In: :disort:LAPACK.f:dger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetf2 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrf In: :disort:LAPACK.f:dgetrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrs In: :disort:LAPACK.f:dgetrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamch In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamc3 Block: dlaswp Block: dscal Block: dswap Block: dtrsm In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: idamax Block: ieeeck Block: ilaenv Block: iparmq In: :disort:LAPACK.f:iparmq get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: lsame Block: scopy Block: sgbtf2 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrf In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrs In: :disort:LAPACK.f:sgbtrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemm In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemv In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sger In: :disort:LAPACK.f:sger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: slaswp Block: stbsv In: :disort:LAPACK.f:stbsv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: strsm In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: xerbla Block: sgbco Block: sgbfa Block: sgbsl Block: sgeco Block: sgefa Block: sgesl Block: sasum Block: saxpy Block: sdot Block: sscal Block: sswap Block: isamax Block: r1mach Block: d1mach Block: i1mach Applying post-processing hooks... character_backward_compatibility_hook Post-processing (stage 2)... Building modules... Building module "disort"... Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "bdref"("bdref")... Constructing wrapper function "bdref"... bdref = bdref(mu,mup,dphi,brdf_type,brdf_arg) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_hapke"... brdf_hapke(mup,mu,dphi,b0,hh,w,pi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rpv"... brdf_rpv(mu_i,mu_r,dphi,rho0,kappa,g_hg,h0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rossli"... brdf_rossli(mu_i,mu_r,dphi,k_iso,k_vol,k_geo,alpha0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "oceabrdf2"... oceabrdf2(do_shadow,refrac_index,ws,mu_i,mu_r,dphi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "shadow_eta"("shadow_eta")... Constructing wrapper function "shadow_eta"... shadow_eta = shadow_eta(cos_theta,sigma_sq,pi) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "disobrdf"... rhoq,rhou,emust,bemst,bdr_beam_analytic = disobrdf(usrang,umu,fbeam,umu0,lamber,albedo,onlyfl,rhoq,rhou,emust,bemst,debug,phi,phi0,bdr_beam_analytic,brdf_type,brdf_arg,nmug,[nstr,numu,nphi]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "surfac2"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' surfac2(albedo,delm0,cmu,fbeam,lamber,mazim,onlyfl,pi,umu,umu0,usrang,bdr,emu,bem,rmu,rhoq,rhou,emust,bemst,debug,gmu,gwt,cosmp,brdf_type,brdf_arg,[mi,mxumu,nn,numu,nazz,nstr,nmug]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "qgausn2"... qgausn2(gmu,gwt,[m]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "zeroit2"... zeroit2(a,[length]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "errmsg"... errmsg(messag,fatal) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtbad"("wrtbad")... Constructing wrapper function "wrtbad"... wrtbad = wrtbad(varnam) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtdim"("wrtdim")... Constructing wrapper function "wrtdim"... wrtdim = wrtdim(dimnam,minval) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "tstbad"("tstbad")... Constructing wrapper function "tstbad"... tstbad = tstbad(varnam,relerr) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgemm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dger"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetf2"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgetf2(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrf"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgetrf(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrs"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgetrs(trans,n,nrhs,a,ipiv,b,info,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamch"("dlamch")... Constructing wrapper function "dlamch"... dlamch = dlamch(cmach) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamc3"("dlamc3")... Constructing wrapper function "dlamc3"... dlamc3 = dlamc3(a,b) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dlaswp"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dlaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dscal"... getarrdims:warning: assumed shape array, using 0 instead of '*' dscal(n,da,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dswap"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dswap(n,dx,incx,dy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dtrsm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dtrsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "idamax"("idamax")... Constructing wrapper function "idamax"... getarrdims:warning: assumed shape array, using 0 instead of '*' idamax = idamax(n,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ieeeck"("ieeeck")... Constructing wrapper function "ieeeck"... ieeeck = ieeeck(ispec,zero,one) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ilaenv"("ilaenv")... Constructing wrapper function "ilaenv"... ilaenv = ilaenv(ispec,name,opts,n1,n2,n3,n4) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '*' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} getarrdims:warning: assumed shape array, using 0 instead of '*' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} iparmq = iparmq(ispec,name,opts,n,ilo,ihi,lwork) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "lsame"("lsame")... Constructing wrapper function "lsame"... lsame = lsame(ca,cb) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "scopy"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' scopy(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtf2"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbtf2(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrf"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbtrf(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrs"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbtrs(trans,n,kl,ku,nrhs,ab,ipiv,b,info,[ldab,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemv"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgemv(trans,m,n,alpha,a,x,incx,beta,y,incy,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sger"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "slaswp"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' slaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "stbsv"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' stbsv(uplo,trans,diag,n,k,a,x,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "strsm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' strsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "xerbla"... xerbla(srname,info) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbco"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbco(abd,n,ml,mu,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbfa"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbfa(abd,n,ml,mu,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbsl"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbsl(abd,n,ml,mu,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgeco"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgeco(a,n,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgefa"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgefa(a,n,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgesl"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgesl(a,n,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sasum"("sasum")... Constructing wrapper function "sasum"... getarrdims:warning: assumed shape array, using 0 instead of '*' sasum = sasum(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "saxpy"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' saxpy(n,sa,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sdot"("sdot")... Constructing wrapper function "sdot"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sdot = sdot(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sscal"... getarrdims:warning: assumed shape array, using 0 instead of '*' sscal(n,sa,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sswap"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sswap(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "isamax"("isamax")... Constructing wrapper function "isamax"... getarrdims:warning: assumed shape array, using 0 instead of '*' isamax = isamax(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "r1mach"("r1mach")... Constructing wrapper function "r1mach"... r1mach = r1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "d1mach"("d1mach")... Constructing wrapper function "d1mach"... d1mach = d1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "i1mach"("i1mach")... Constructing wrapper function "i1mach"... i1mach = i1mach(i) Wrote C/API module "disort" to file "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c" Fortran 77 wrappers are saved to "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f" INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c' to sources. INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9' to include_dirs. copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.c -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.h -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f' to sources. INFO: build_src: building npy-pkg config files running build_ext INFO: customize UnixCCompiler INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmp27l9smg9/tmp creating /tmp/tmp27l9smg9/tmp/tmp27l9smg9 INFO: compile options: '-MMD -MF /tmp/tmp27l9smg9/file.c.d -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmp27l9smg9/file.c INFO: customize UnixCCompiler using build_ext INFO: get_default_fcompiler: matching types: '['arm', 'gnu95', 'intel', 'lahey', 'pg', 'nv', 'absoft', 'nag', 'vast', 'compaq', 'intele', 'intelem', 'gnu', 'g95', 'pathf95', 'nagfor', 'fujitsu']' INFO: customize ArmFlangCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-gfortran WARN: Could not locate executable armflang INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ar INFO: customize Gnu95FCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ld INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ranlib INFO: customize Gnu95FCompiler INFO: customize Gnu95FCompiler using build_ext INFO: building 'disort' extension INFO: compiling C sources INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmpuul0531u/tmp creating /tmp/tmpuul0531u/tmp/tmpuul0531u creating /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: compile options: '-DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 | capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); | ^ /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 | capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); | ^ error: Command "/home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC -DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c -o /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o -MMD -MF /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o.d" failed with exit status 1 ``` Where it seems that the relevant error is from changes to character handling: ```bash getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 | capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); | ^ /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 | capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); | ``` OTOH I would suspect something in the character handling of https://github.com/numpy/numpy/issues/23356 or https://github.com/numpy/numpy/pull/23194 Diffing the outputs b/w `1.24.3` and `1.24.4` only show: ```bash Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' ``` Indeed, `f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f skip: iparmq` does seem to compile as well. Will investigate ASAP. P.S. @kconnour the code being tested is [Pythonic-Distort](https://github.com/LDEO-CREW/Pythonic-DISORT/tree/main/disort4.0.99_f2py) right? Hi @HaoZeke, thanks for investigating! I'm really glad to see someone else is getting a similar error. Actually, the code I'm testing is on the api branch of [my repo](https://github.com/kconnour/pyRT_DISORT) but it should be extremely similar to the code in the repo you linked. We're both apparently trying to make a front-end to a popular open-source fortran algorithm. Note that I coded a workaround to this problem in pyproject.toml, where I force it to install numpy==1.24.0 in order to circumvent this issue... so if you ran the installation script from my repo, it shouldn't encounter this error. I'm happy to provide any additional info to help diagnose this issue! We ran into similar issues with our code, and after testing, it's because of having a decimal inside of a parameter declaration. Once that happens, everything later stops processing properly. The same seems be happening in the examples above If you look at the error log posted, https://github.com/numpy/numpy/issues/24008#issuecomment-1601586519, you'll see the first error message is the attached. ``` Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 ``` You could confirm whether this also works for your code by modifying that parameter to just be ONE=1, ZERO=0, and recompiling. If it gets past that point, that's the problem. We worked around this by reverting crackfortran.py to prior to the changes introduced here: https://github.com/numpy/numpy/pull/23637/files As to speculation: I think that the changes to the if blocks in https://github.com/numpy/numpy/pull/23637/commits caused an issue that assumes the input is an integer. Even more speculative - I see the kind selector is in that portion, and the kind selector requires integers. (https://numpy.org/doc/stable/f2py/advanced.html#dealing-with-kind-specifiers) . But I'm not sure - I didn't spend enough time on understanding crackfortran.py once we discovered that reverting crackfortran.py fixed our problem.

2023-07-31T06:34:32Z

[]

Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error

10,207

numpy/numpy

numpy__numpy-24511

688f86133c78a3be569981a121d77166527d5b38

diff --git a/numpy/_build_utils/gitversion.py b/numpy/_build_utils/gitversion.py --- a/numpy/_build_utils/gitversion.py +++ b/numpy/_build_utils/gitversion.py @@ -24,6 +24,7 @@ def git_version(version): import subprocess import os.path + git_hash = '' try: p = subprocess.Popen( ['git', 'log', '-1', '--format="%H %aI"'], @@ -48,8 +49,6 @@ def git_version(version): # Only attach git tag to development versions if 'dev' in version: version += f'+git{git_date}.{git_hash[:7]}' - else: - git_hash = '' return version, git_hash

BUG: numpy 1.26.0b1 fails to build from sdist when no git is present ### Describe the issue: ---- `<edit>`: (@mattip) Adding the root cause: The version is missing from the sdist for 1.26b1, resulting in an attempt to get it via `git`. But that does not work because ... `<edit>` ---- The `gitversion.py` script introduced in #24196 is broken when no `git` binary is available. ### Reproduce the code example: ```python $ python3 numpy/_build_utils/gitversion.py ``` ### Error message: ```shell $ python3 numpy/_build_utils/gitversion.py Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value ``` ### Runtime information: N/A ### Context for the issue: https://github.com/sagemath/sage/pull/36123

Thanks for the report @mkoeppe. That's a bug indeed. The `gitversion.py` script should check for a file containing the needed git hash (as was done with `if fs.exists('_version_meson.py')` before gh-24196). Building from an sdist should work when `git` is not installed. @stefanv can you please have a look at this?

2023-08-23T13:02:33Z

[]

Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value

10,212

numpy/numpy

numpy__numpy-24522

ef94040e9a7db18620d825782b8a14650f7c3346

diff --git a/numpy/_build_utils/gitversion.py b/numpy/_build_utils/gitversion.py --- a/numpy/_build_utils/gitversion.py +++ b/numpy/_build_utils/gitversion.py @@ -24,6 +24,7 @@ def git_version(version): import subprocess import os.path + git_hash = '' try: p = subprocess.Popen( ['git', 'log', '-1', '--format="%H %aI"'], @@ -48,8 +49,6 @@ def git_version(version): # Only attach git tag to development versions if 'dev' in version: version += f'+git{git_date}.{git_hash[:7]}' - else: - git_hash = '' return version, git_hash

BUG: numpy 1.26.0b1 fails to build from sdist when no git is present ### Describe the issue: ---- `<edit>`: (@mattip) Adding the root cause: The version is missing from the sdist for 1.26b1, resulting in an attempt to get it via `git`. But that does not work because ... `<edit>` ---- The `gitversion.py` script introduced in #24196 is broken when no `git` binary is available. ### Reproduce the code example: ```python $ python3 numpy/_build_utils/gitversion.py ``` ### Error message: ```shell $ python3 numpy/_build_utils/gitversion.py Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value ``` ### Runtime information: N/A ### Context for the issue: https://github.com/sagemath/sage/pull/36123

Thanks for the report @mkoeppe. That's a bug indeed. The `gitversion.py` script should check for a file containing the needed git hash (as was done with `if fs.exists('_version_meson.py')` before gh-24196). Building from an sdist should work when `git` is not installed. @stefanv can you please have a look at this?

2023-08-24T16:53:29Z

[]

Traceback (most recent call last): File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 68, in <module> version, git_hash = git_version(init_version()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/sage/local/var/lib/sage/venv-python3.11/var/tmp/sage/build/numpy-1.26.0b1/src/numpy/_build_utils/gitversion.py", line 53, in git_version return version, git_hash ^^^^^^^^ UnboundLocalError: cannot access local variable 'git_hash' where it is not associated with a value

10,213

numpy/numpy

numpy__numpy-24542

9d64cc2bbe44292c46af768fb5da9b7968c2c6cb

diff --git a/numpy/f2py/crackfortran.py b/numpy/f2py/crackfortran.py --- a/numpy/f2py/crackfortran.py +++ b/numpy/f2py/crackfortran.py @@ -1742,6 +1742,23 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: del d1[k] + if 'len' in d1 and 'array' in d1: + if d1['len'] == '': + d1['len'] = d1['array'] + del d1['array'] + elif typespec == 'character': + if ('charselector' not in edecl) or (not edecl['charselector']): + edecl['charselector'] = {} + if 'len' in edecl['charselector']: + del edecl['charselector']['len'] + edecl['charselector']['*'] = d1['len'] + del d1['len'] + else: + d1['array'] = d1['array'] + ',' + d1['len'] + del d1['len'] + errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( + typespec, e, typespec, ename, d1['array'])) + if 'len' in d1: if typespec in ['complex', 'integer', 'logical', 'real']: if ('kindselector' not in edecl) or (not edecl['kindselector']): @@ -1763,16 +1780,6 @@ def updatevars(typespec, selector, attrspec, entitydecl): else: edecl['='] = d1['init'] - if 'len' in d1 and 'array' in d1: - if d1['len'] == '': - d1['len'] = d1['array'] - del d1['array'] - else: - d1['array'] = d1['array'] + ',' + d1['len'] - del d1['len'] - errmess('updatevars: "%s %s" is mapped to "%s %s(%s)"\n' % ( - typespec, e, typespec, ename, d1['array'])) - if 'array' in d1: dm = 'dimension(%s)' % d1['array'] if 'attrspec' not in edecl or (not edecl['attrspec']):

BUG: f2py cannot compile files it used to be able to compile ### Describe the issue: I'm attempting to use f2py to compile some fortran code. I'm able to do this using the following numpy versions: - 1.21.0 - 1.22.0 - 1.23.0 - 1.24.0 - 1.24.2 Starting on v1.24.3, the same compilation code no longer works using f2py. Strangely enough, if I precompile one of files (LAPACK.f), the compilation can work with v1.24.3. ### Reproduce the code example: ```python # I don't see how I can produce runnable code using f2py, as it requires the fortran source code # This code produces a .so file in my project's home directory on numpy <= 1.24.2 but doesn't work on numpy >= 1.24.3 from pathlib import Path from numpy import f2py project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.f', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) # If I precompile LAPACK.f using: # /usr/bin/gfortran -Wall -g -ffixed-form -fno-second-underscore -g -fno-second-underscore -fPIC -O3 -funroll-loops -c LAPACK.f # then it works using numpy = 1.24.3. Note the only difference in the code is the LAPACK.f is now LAPACK.o project_path = Path(__file__).resolve().parent disort_directory = project_path.joinpath('disort4.0.99') module_name = 'disort' fortran_source_filenames = ['BDREF.f', 'DISOBRDF.f', 'ERRPACK.f', 'LAPACK.o', 'LINPAK.f', 'RDI1MACH.f'] fortran_paths = [disort_directory.joinpath(f) for f in fortran_source_filenames] with open(disort_directory.joinpath('DISORT.f')) as disort_module: f2py.compile(disort_module.read(), modulename=module_name, extra_args=fortran_paths) ``` ### Error message: ```shell There is no error message. f2py simply stops right before where it prints this line: INFO: compiling Fortran sources ``` ### Runtime information: Line 1 output: 1.24.0 3.10.6 (main, May 29 2023, 11:10:38) [GCC 11.3.0] (note that I have no idea why it says 1.24.0 when Pycharm assures me I'm using 1.24.2) Line 2 output: Exception ignored on calling ctypes callback function: <function ThreadpoolController._find_libraries_with_dl_iterate_phdr.<locals>.match_library_callback at 0x7f592f9c00d0> Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error [{'simd_extensions': {'baseline': ['SSE', 'SSE2', 'SSE3'], 'found': ['SSSE3', 'SSE41', 'POPCNT', 'SSE42', 'AVX', 'F16C', 'FMA3', 'AVX2'], 'not_found': ['AVX512F', 'AVX512CD', 'AVX512_KNL', 'AVX512_KNM', 'AVX512_SKX', 'AVX512_CLX', 'AVX512_CNL', 'AVX512_ICL']}}] None ### Context for the issue: I believe that LAPACK.f is one of the more widely used fortran codes. If it cannot compile in conjunction with other code, that could potentially disrupt a good number of users.

Ping @HaoZeke, also to confirm, can you check if the issue persists on 1.25.0? @seberg I can confirm that the issue persists on 1.25.0. It's actually why I noticed it in the first place. My code that used to run failed and then I tracked it down to the version number described above, in hopes that someone might have a better idea what caused it. On `1.24.4` the error is: ```bash ❯ f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f running build running config_cc INFO: unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc INFO: unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src INFO: build_src INFO: building extension "disort" sources INFO: f2py options: [] INFO: f2py:> /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c creating /tmp/tmpuul0531u/src.linux-x86_64-3.9 Reading fortran codes... Reading file 'BDREF.f' (format:fix,strict) Reading file 'DISOBRDF.f' (format:fix,strict) rmbadname1: Replacing "float" with "float_bn". rmbadname1: Replacing "len" with "len_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "float" with "float_bn". Reading file 'ERRPACK.f' (format:fix,strict) Reading file 'LAPACK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #745 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #1363 in LAPACK.f:" PARAMETER ( ONE = 1.0D+0, ZERO = 0.0D+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". Line #2115 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "char" with "char_bn". rmbadname1: Replacing "int" with "int_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #3999 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". Line #4273 in LAPACK.f:" PARAMETER ( ONE = 1.0E+0, ZERO = 0.0E+0 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #4277 in LAPACK.f:" PARAMETER ( NBMAX = 64, LDWORK = NBMAX+1 )" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #5124 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Line #5462 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Line #6657 in LAPACK.f:" PARAMETER (ONE=1.0E+0,ZERO=0.0E+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Reading file 'LINPAK.f' (format:fix,strict) rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "min" with "min_bn". rmbadname1: Replacing "max" with "max_bn". Reading file 'RDI1MACH.f' (format:fix,strict) Post-processing... Block: disort Block: bdref Block: brdf_hapke Block: brdf_rpv Block: brdf_rossli Block: oceabrdf2 Block: shadow_eta Block: disobrdf {} In: :disort:DISOBRDF.f:surfac2 vars2fortran: No typespec for argument "nazz". Block: surfac2 Block: qgausn2 Block: zeroit2 Block: errmsg Block: wrtbad Block: wrtdim Block: tstbad Block: dgemm In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dger In: :disort:LAPACK.f:dger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetf2 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dgetf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrf In: :disort:LAPACK.f:dgetrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dgetrs In: :disort:LAPACK.f:dgetrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamch In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dlamch get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: dlamc3 Block: dlaswp Block: dscal Block: dswap Block: dtrsm In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:dtrsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: idamax Block: ieeeck Block: ilaenv Block: iparmq In: :disort:LAPACK.f:iparmq get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: lsame Block: scopy Block: sgbtf2 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtf2 get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrf In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgbtrf get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgbtrs In: :disort:LAPACK.f:sgbtrs get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemm In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sgemv In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:sgemv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: sger In: :disort:LAPACK.f:sger get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: slaswp Block: stbsv In: :disort:LAPACK.f:stbsv get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: strsm In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 In: :disort:LAPACK.f:strsm get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 Block: xerbla Block: sgbco Block: sgbfa Block: sgbsl Block: sgeco Block: sgefa Block: sgesl Block: sasum Block: saxpy Block: sdot Block: sscal Block: sswap Block: isamax Block: r1mach Block: d1mach Block: i1mach Applying post-processing hooks... character_backward_compatibility_hook Post-processing (stage 2)... Building modules... Building module "disort"... Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "bdref"("bdref")... Constructing wrapper function "bdref"... bdref = bdref(mu,mup,dphi,brdf_type,brdf_arg) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_hapke"... brdf_hapke(mup,mu,dphi,b0,hh,w,pi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rpv"... brdf_rpv(mu_i,mu_r,dphi,rho0,kappa,g_hg,h0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "brdf_rossli"... brdf_rossli(mu_i,mu_r,dphi,k_iso,k_vol,k_geo,alpha0,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "oceabrdf2"... oceabrdf2(do_shadow,refrac_index,ws,mu_i,mu_r,dphi,brdf) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "shadow_eta"("shadow_eta")... Constructing wrapper function "shadow_eta"... shadow_eta = shadow_eta(cos_theta,sigma_sq,pi) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "disobrdf"... rhoq,rhou,emust,bemst,bdr_beam_analytic = disobrdf(usrang,umu,fbeam,umu0,lamber,albedo,onlyfl,rhoq,rhou,emust,bemst,debug,phi,phi0,bdr_beam_analytic,brdf_type,brdf_arg,nmug,[nstr,numu,nphi]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "surfac2"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' surfac2(albedo,delm0,cmu,fbeam,lamber,mazim,onlyfl,pi,umu,umu0,usrang,bdr,emu,bem,rmu,rhoq,rhou,emust,bemst,debug,gmu,gwt,cosmp,brdf_type,brdf_arg,[mi,mxumu,nn,numu,nazz,nstr,nmug]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "qgausn2"... qgausn2(gmu,gwt,[m]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "zeroit2"... zeroit2(a,[length]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "errmsg"... errmsg(messag,fatal) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtbad"("wrtbad")... Constructing wrapper function "wrtbad"... wrtbad = wrtbad(varnam) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "wrtdim"("wrtdim")... Constructing wrapper function "wrtdim"... wrtdim = wrtdim(dimnam,minval) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "tstbad"("tstbad")... Constructing wrapper function "tstbad"... tstbad = tstbad(varnam,relerr) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgemm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dger"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetf2"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgetf2(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrf"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgetrf(m,n,a,ipiv,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dgetrs"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dgetrs(trans,n,nrhs,a,ipiv,b,info,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamch"("dlamch")... Constructing wrapper function "dlamch"... dlamch = dlamch(cmach) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "dlamc3"("dlamc3")... Constructing wrapper function "dlamc3"... dlamc3 = dlamc3(a,b) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dlaswp"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dlaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dscal"... getarrdims:warning: assumed shape array, using 0 instead of '*' dscal(n,da,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dswap"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dswap(n,dx,incx,dy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "dtrsm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' dtrsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "idamax"("idamax")... Constructing wrapper function "idamax"... getarrdims:warning: assumed shape array, using 0 instead of '*' idamax = idamax(n,dx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ieeeck"("ieeeck")... Constructing wrapper function "ieeeck"... ieeeck = ieeeck(ispec,zero,one) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "ilaenv"("ilaenv")... Constructing wrapper function "ilaenv"... ilaenv = ilaenv(ispec,name,opts,n1,n2,n3,n4) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '*' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} getarrdims:warning: assumed shape array, using 0 instead of '*' getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} iparmq = iparmq(ispec,name,opts,n,ilo,ihi,lwork) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "lsame"("lsame")... Constructing wrapper function "lsame"... lsame = lsame(ca,cb) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "scopy"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' scopy(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtf2"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbtf2(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrf"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbtrf(m,n,kl,ku,ab,ipiv,info,[ldab]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbtrs"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbtrs(trans,n,kl,ku,nrhs,ab,ipiv,b,info,[ldab,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgemm(transa,transb,m,n,k,alpha,a,b,beta,c,[lda,ldb,ldc]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgemv"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgemv(trans,m,n,alpha,a,x,incx,beta,y,incy,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sger"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sger(m,n,alpha,x,incx,y,incy,a,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "slaswp"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' slaswp(n,a,k1,k2,ipiv,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "stbsv"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' stbsv(uplo,trans,diag,n,k,a,x,incx,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "strsm"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' strsm(side,uplo,transa,diag,m,n,alpha,a,b,[lda,ldb]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "xerbla"... xerbla(srname,info) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbco"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbco(abd,n,ml,mu,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbfa"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbfa(abd,n,ml,mu,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgbsl"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgbsl(abd,n,ml,mu,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgeco"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgeco(a,n,ipvt,rcond,z,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgefa"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgefa(a,n,ipvt,info,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sgesl"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sgesl(a,n,ipvt,b,job,[lda]) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sasum"("sasum")... Constructing wrapper function "sasum"... getarrdims:warning: assumed shape array, using 0 instead of '*' sasum = sasum(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "saxpy"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' saxpy(n,sa,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "sdot"("sdot")... Constructing wrapper function "sdot"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sdot = sdot(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sscal"... getarrdims:warning: assumed shape array, using 0 instead of '*' sscal(n,sa,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Constructing wrapper function "sswap"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' sswap(n,sx,incx,sy,incy) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "isamax"("isamax")... Constructing wrapper function "isamax"... getarrdims:warning: assumed shape array, using 0 instead of '*' isamax = isamax(n,sx,incx) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "r1mach"("r1mach")... Constructing wrapper function "r1mach"... r1mach = r1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "d1mach"("d1mach")... Constructing wrapper function "d1mach"... d1mach = d1mach(i) Generating possibly empty wrappers" Maybe empty "disort-f2pywrappers.f" Creating wrapper for Fortran function "i1mach"("i1mach")... Constructing wrapper function "i1mach"... i1mach = i1mach(i) Wrote C/API module "disort" to file "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c" Fortran 77 wrappers are saved to "/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f" INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c' to sources. INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9' to include_dirs. copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.c -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 copying /home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/f2py/src/fortranobject.h -> /tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: adding '/tmp/tmpuul0531u/src.linux-x86_64-3.9/disort-f2pywrappers.f' to sources. INFO: build_src: building npy-pkg config files running build_ext INFO: customize UnixCCompiler INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmp27l9smg9/tmp creating /tmp/tmp27l9smg9/tmp/tmp27l9smg9 INFO: compile options: '-MMD -MF /tmp/tmp27l9smg9/file.c.d -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmp27l9smg9/file.c INFO: customize UnixCCompiler using build_ext INFO: get_default_fcompiler: matching types: '['arm', 'gnu95', 'intel', 'lahey', 'pg', 'nv', 'absoft', 'nag', 'vast', 'compaq', 'intele', 'intelem', 'gnu', 'g95', 'pathf95', 'nagfor', 'fujitsu']' INFO: customize ArmFlangCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-gfortran WARN: Could not locate executable armflang INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ar INFO: customize Gnu95FCompiler INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ld INFO: Found executable /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-ranlib INFO: customize Gnu95FCompiler INFO: customize Gnu95FCompiler using build_ext INFO: building 'disort' extension INFO: compiling C sources INFO: C compiler: /home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC creating /tmp/tmpuul0531u/tmp creating /tmp/tmpuul0531u/tmp/tmpuul0531u creating /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9 INFO: compile options: '-DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c' INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c INFO: x86_64-conda-linux-gnu-cc: /tmp/tmpuul0531u/src.linux-x86_64-3.9/fortranobject.c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 | capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); | ^ /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 | capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); | ^ error: Command "/home/rgoswami/.micromamba/envs/numpy-dev/bin/x86_64-conda-linux-gnu-cc -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -march=nocona -mtune=haswell -ftree-vectorize -fPIC -fstack-protector-strong -fno-plt -O2 -ffunction-sections -pipe -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -DNDEBUG -D_FORTIFY_SOURCE=2 -O2 -isystem /home/rgoswami/.micromamba/envs/numpy-dev/include -fPIC -DNPY_DISABLE_OPTIMIZATION=1 -I/tmp/tmpuul0531u/src.linux-x86_64-3.9 -I/home/rgoswami/Git/Github/Quansight/f2py_envs/numpy/numpy/core/include -I/home/rgoswami/.micromamba/envs/numpy-dev/include/python3.9 -c /tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.c -o /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o -MMD -MF /tmp/tmpuul0531u/tmp/tmpuul0531u/src.linux-x86_64-3.9/disortmodule.o.d" failed with exit status 1 ``` Where it seems that the relevant error is from changes to character handling: ```bash getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} getstrlength: expected a signature of a string but got: {'typespec': 'character', 'charselector': {'*': ''}, 'attrspec': [], 'dimension': ['*']} /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c: In function 'f2py_rout_disort_iparmq': /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4772:58: error: expected expression before ',' token 4772 | capi_name_as_array = ndarray_from_pyobj( NPY_STRING,,name_Dims,name_Rank, capi_name_intent,name_capi,capi_errmess); | ^ /tmp/tmp5l8spqw9/src.linux-x86_64-3.9/disortmodule.c:4787:58: error: expected expression before ',' token 4787 | capi_opts_as_array = ndarray_from_pyobj( NPY_STRING,,opts_Dims,opts_Rank, capi_opts_intent,opts_capi,capi_errmess); | ``` OTOH I would suspect something in the character handling of https://github.com/numpy/numpy/issues/23356 or https://github.com/numpy/numpy/pull/23194 Diffing the outputs b/w `1.24.3` and `1.24.4` only show: ```bash Creating wrapper for Fortran function "iparmq"("iparmq")... Constructing wrapper function "iparmq"... getarrdims:warning: assumed shape array, using 0 instead of '*' getarrdims:warning: assumed shape array, using 0 instead of '*' ``` Indeed, `f2py -c --f90flags='-O3' -m disort BDREF.f DISOBRDF.f ERRPACK.f LAPACK.f LINPAK.f RDI1MACH.f skip: iparmq` does seem to compile as well. Will investigate ASAP. P.S. @kconnour the code being tested is [Pythonic-Distort](https://github.com/LDEO-CREW/Pythonic-DISORT/tree/main/disort4.0.99_f2py) right? Hi @HaoZeke, thanks for investigating! I'm really glad to see someone else is getting a similar error. Actually, the code I'm testing is on the api branch of [my repo](https://github.com/kconnour/pyRT_DISORT) but it should be extremely similar to the code in the repo you linked. We're both apparently trying to make a front-end to a popular open-source fortran algorithm. Note that I coded a workaround to this problem in pyproject.toml, where I force it to install numpy==1.24.0 in order to circumvent this issue... so if you ran the installation script from my repo, it shouldn't encounter this error. I'm happy to provide any additional info to help diagnose this issue! We ran into similar issues with our code, and after testing, it's because of having a decimal inside of a parameter declaration. Once that happens, everything later stops processing properly. The same seems be happening in the examples above If you look at the error log posted, https://github.com/numpy/numpy/issues/24008#issuecomment-1601586519, you'll see the first error message is the attached. ``` Line #229 in LAPACK.f:" PARAMETER (ONE=1.0D+0,ZERO=0.0D+0)" get_parameters: got "eval() arg 1 must be a string, bytes or code object" on 4 ``` You could confirm whether this also works for your code by modifying that parameter to just be ONE=1, ZERO=0, and recompiling. If it gets past that point, that's the problem. We worked around this by reverting crackfortran.py to prior to the changes introduced here: https://github.com/numpy/numpy/pull/23637/files As to speculation: I think that the changes to the if blocks in https://github.com/numpy/numpy/pull/23637/commits caused an issue that assumes the input is an integer. Even more speculative - I see the kind selector is in that portion, and the kind selector requires integers. (https://numpy.org/doc/stable/f2py/advanced.html#dealing-with-kind-specifiers) . But I'm not sure - I didn't spend enough time on understanding crackfortran.py once we discovered that reverting crackfortran.py fixed our problem.

2023-08-25T14:39:05Z

[]

Traceback (most recent call last): File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 584, in match_library_callback self._make_controller_from_path(filepath) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 725, in _make_controller_from_path lib_controller = lib_controller_class( File "~/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 842, in __init__ super().__init__(**kwargs) File "~/repos/pyRT_DISORT/venv/lib/python3.10/site-packages/threadpoolctl.py", line 810, in __init__ self._dynlib = ctypes.CDLL(filepath, mode=_RTLD_NOLOAD) File "/usr/lib/python3.10/ctypes/__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) OSError: dlopen() error

10,215

numpy/numpy

numpy__numpy-258

b452014f0f9e8e6a69c6b95a62d70e3d99b9c0f9

diff --git a/numpy/distutils/ccompiler.py b/numpy/distutils/ccompiler.py --- a/numpy/distutils/ccompiler.py +++ b/numpy/distutils/ccompiler.py @@ -58,7 +58,11 @@ def CCompiler_spawn(self, cmd, display=None): if s: if is_sequence(cmd): cmd = ' '.join(list(cmd)) - print(o) + try: + print(o) + except UnicodeError: + # When installing through pip, `o` can contain non-ascii chars + pass if re.search('Too many open files', o): msg = '\nTry rerunning setup command until build succeeds.' else: diff --git a/numpy/distutils/command/build_clib.py b/numpy/distutils/command/build_clib.py --- a/numpy/distutils/command/build_clib.py +++ b/numpy/distutils/command/build_clib.py @@ -81,21 +81,23 @@ def run(self): if self.have_f_sources(): from numpy.distutils.fcompiler import new_fcompiler - self.fcompiler = new_fcompiler(compiler=self.fcompiler, - verbose=self.verbose, - dry_run=self.dry_run, - force=self.force, - requiref90='f90' in languages, - c_compiler=self.compiler) - if self.fcompiler is not None: - self.fcompiler.customize(self.distribution) + self._f_compiler = new_fcompiler(compiler=self.fcompiler, + verbose=self.verbose, + dry_run=self.dry_run, + force=self.force, + requiref90='f90' in languages, + c_compiler=self.compiler) + if self._f_compiler is not None: + self._f_compiler.customize(self.distribution) libraries = self.libraries self.libraries = None - self.fcompiler.customize_cmd(self) + self._f_compiler.customize_cmd(self) self.libraries = libraries - self.fcompiler.show_customization() + self._f_compiler.show_customization() + else: + self._f_compiler = None self.build_libraries(self.libraries) @@ -121,7 +123,7 @@ def build_libraries(self, libraries): def build_a_library(self, build_info, lib_name, libraries): # default compilers compiler = self.compiler - fcompiler = self.fcompiler + fcompiler = self._f_compiler sources = build_info.get('sources') if sources is None or not is_sequence(sources): @@ -233,7 +235,7 @@ def build_a_library(self, build_info, lib_name, libraries): debug=self.debug, extra_postargs=extra_postargs) - if requiref90 and self.fcompiler.module_dir_switch is None: + if requiref90 and self._f_compiler.module_dir_switch is None: # move new compiled F90 module files to module_build_dir for f in glob('*.mod'): if f in existing_modules: diff --git a/numpy/distutils/fcompiler/hpux.py b/numpy/distutils/fcompiler/hpux.py --- a/numpy/distutils/fcompiler/hpux.py +++ b/numpy/distutils/fcompiler/hpux.py @@ -9,17 +9,17 @@ class HPUXFCompiler(FCompiler): version_pattern = r'HP F90 (?P<version>[^\s*,]*)' executables = { - 'version_cmd' : ["<F90>", "+version"], + 'version_cmd' : ["f90", "+version"], 'compiler_f77' : ["f90"], 'compiler_fix' : ["f90"], 'compiler_f90' : ["f90"], - 'linker_so' : None, + 'linker_so' : ["ld", "-b"], 'archiver' : ["ar", "-cr"], 'ranlib' : ["ranlib"] } module_dir_switch = None #XXX: fix me module_include_switch = None #XXX: fix me - pic_flags = ['+pic=long'] + pic_flags = ['+Z'] def get_flags(self): return self.pic_flags + ['+ppu', '+DD64'] def get_flags_opt(self): diff --git a/numpy/distutils/fcompiler/ibm.py b/numpy/distutils/fcompiler/ibm.py --- a/numpy/distutils/fcompiler/ibm.py +++ b/numpy/distutils/fcompiler/ibm.py @@ -12,7 +12,7 @@ class IBMFCompiler(FCompiler): compiler_type = 'ibm' description = 'IBM XL Fortran Compiler' - version_pattern = r'(xlf$1$\s*|)IBM XL Fortran ((Advanced Edition |)Version |Enterprise Edition V)(?P<version>[^\s*]*)' + version_pattern = r'(xlf$1$\s*|)IBM XL Fortran ((Advanced Edition |)Version |Enterprise Edition V|for AIX, V)(?P<version>[^\s*]*)' #IBM XL Fortran Enterprise Edition V10.1 for AIX \nVersion: 10.01.0000.0004 executables = { @@ -86,7 +86,7 @@ def get_flags_linker_so(self): return opt def get_flags_opt(self): - return ['-O5'] + return ['-O3'] if __name__ == '__main__': log.set_verbosity(2) diff --git a/numpy/distutils/fcompiler/pg.py b/numpy/distutils/fcompiler/pg.py --- a/numpy/distutils/fcompiler/pg.py +++ b/numpy/distutils/fcompiler/pg.py @@ -10,14 +10,14 @@ class PGroupFCompiler(FCompiler): compiler_type = 'pg' description = 'Portland Group Fortran Compiler' - version_pattern = r'\s*pg(f77|f90|hpf) (?P<version>[\d.-]+).*' + version_pattern = r'\s*pg(f77|f90|hpf|fortran) (?P<version>[\d.-]+).*' if platform == 'darwin': executables = { - 'version_cmd' : ["<F77>", "-V 2>/dev/null"], - 'compiler_f77' : ["pgf77", "-dynamiclib"], - 'compiler_fix' : ["pgf90", "-Mfixed", "-dynamiclib"], - 'compiler_f90' : ["pgf90", "-dynamiclib"], + 'version_cmd' : ["<F77>", "-V"], + 'compiler_f77' : ["pgfortran", "-dynamiclib"], + 'compiler_fix' : ["pgfortran", "-Mfixed", "-dynamiclib"], + 'compiler_f90' : ["pgfortran", "-dynamiclib"], 'linker_so' : ["libtool"], 'archiver' : ["ar", "-cr"], 'ranlib' : ["ranlib"] @@ -25,11 +25,11 @@ class PGroupFCompiler(FCompiler): pic_flags = [''] else: executables = { - 'version_cmd' : ["<F77>", "-V 2>/dev/null"], - 'compiler_f77' : ["pgf77"], - 'compiler_fix' : ["pgf90", "-Mfixed"], - 'compiler_f90' : ["pgf90"], - 'linker_so' : ["pgf90","-shared","-fpic"], + 'version_cmd' : ["<F77>", "-V"], + 'compiler_f77' : ["pgfortran"], + 'compiler_fix' : ["pgfortran", "-Mfixed"], + 'compiler_f90' : ["pgfortran"], + 'linker_so' : ["pgfortran","-shared","-fpic"], 'archiver' : ["ar", "-cr"], 'ranlib' : ["ranlib"] } diff --git a/numpy/distutils/system_info.py b/numpy/distutils/system_info.py --- a/numpy/distutils/system_info.py +++ b/numpy/distutils/system_info.py @@ -125,6 +125,7 @@ from distutils.dist import Distribution import distutils.sysconfig from distutils import log +from distutils.util import get_platform from numpy.distutils.exec_command import \ find_executable, exec_command, get_pythonexe @@ -193,14 +194,23 @@ def libpaths(paths,bits): '/opt/local/lib','/sw/lib'], platform_bits) default_include_dirs = ['/usr/local/include', '/opt/include', '/usr/include', + # path of umfpack under macports + '/opt/local/include/ufsparse', '/opt/local/include', '/sw/include', '/usr/include/suitesparse'] default_src_dirs = ['.','/usr/local/src', '/opt/src','/sw/src'] - default_x11_lib_dirs = libpaths(['/usr/X11R6/lib','/usr/X11/lib', '/usr/lib'], platform_bits) default_x11_include_dirs = ['/usr/X11R6/include','/usr/X11/include', '/usr/include'] + if os.path.exists('/usr/lib/X11'): + globbed_x11_dir = glob('/usr/lib/*/libX11.so') + if globbed_x11_dir: + x11_so_dir = os.path.split(globbed_x11_dir[0])[0] + default_x11_lib_dirs.extend([x11_so_dir, '/usr/lib/X11']) + default_x11_include_dirs.extend(['/usr/lib/X11/include', + '/usr/include/X11']) + if os.path.join(sys.prefix, 'lib') not in default_lib_dirs: default_lib_dirs.insert(0,os.path.join(sys.prefix, 'lib')) @@ -1273,7 +1283,6 @@ def get_atlas_version(**config): result = _cached_atlas_version[key] = atlas_version, info return result -from distutils.util import get_platform class lapack_opt_info(system_info): @@ -1284,7 +1293,8 @@ def calc_info(self): if sys.platform=='darwin' and not os.environ.get('ATLAS',None): args = [] link_args = [] - if get_platform()[-4:] == 'i386': + if get_platform()[-4:] == 'i386' or 'intel' in get_platform() or \ + 'i386' in platform.platform(): intel = 1 else: intel = 0 @@ -1371,7 +1381,8 @@ def calc_info(self): if sys.platform=='darwin' and not os.environ.get('ATLAS',None): args = [] link_args = [] - if get_platform()[-4:] == 'i386': + if get_platform()[-4:] == 'i386' or 'intel' in get_platform() or \ + 'i386' in platform.platform(): intel = 1 else: intel = 0 diff --git a/numpy/distutils/unixccompiler.py b/numpy/distutils/unixccompiler.py --- a/numpy/distutils/unixccompiler.py +++ b/numpy/distutils/unixccompiler.py @@ -17,6 +17,18 @@ # Note that UnixCCompiler._compile appeared in Python 2.3 def UnixCCompiler__compile(self, obj, src, ext, cc_args, extra_postargs, pp_opts): """Compile a single source files with a Unix-style compiler.""" + # HP ad-hoc fix, see ticket 1383 + ccomp = self.compiler_so + if ccomp[0] == 'aCC': + # remove flags that will trigger ANSI-C mode for aCC + if '-Ae' in ccomp: + ccomp.remove('-Ae') + if '-Aa' in ccomp: + ccomp.remove('-Aa') + # add flags for (almost) sane C++ handling + ccomp += ['-AA'] + self.compiler_so = ccomp + display = '%s: %s' % (os.path.basename(self.compiler_so[0]),src) try: self.spawn(self.compiler_so + cc_args + [src, '-o', obj] + diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -114,7 +114,12 @@ def write_version_py(filename='numpy/version.py'): GIT_REVISION = git_version() elif os.path.exists('numpy/version.py'): # must be a source distribution, use existing version file - from numpy.version import git_revision as GIT_REVISION + try: + from numpy.version import git_revision as GIT_REVISION + except ImportError: + raise ImportError("Unable to import git_revision. Try removing " \ + "numpy/version.py and the build directory " \ + "before building.") else: GIT_REVISION = "Unknown" @@ -162,6 +167,19 @@ def setup_package(): if os.path.isfile(site_cfg): shutil.copy(site_cfg, src_path) + # Ugly hack to make pip work with Python 3, see #1857. + # Explanation: pip messes with __file__ which interacts badly with the + # change in directory due to the 2to3 conversion. Therefore we restore + # __file__ to what it would have been otherwise. + global __file__ + __file__ = os.path.join(os.curdir, os.path.basename(__file__)) + if '--egg-base' in sys.argv: + # Change pip-egg-info entry to absolute path, so pip can find it + # after changing directory. + idx = sys.argv.index('--egg-base') + if sys.argv[idx + 1] == 'pip-egg-info': + sys.argv[idx + 1] = os.path.join(local_path, 'pip-egg-info') + old_path = os.getcwd() os.chdir(src_path) sys.path.insert(0, src_path)

Error in linalg.norm() (Trac #785) _Original ticket http://projects.scipy.org/numpy/ticket/785 on 2008-05-09 by trac user nick, assigned to unknown._ While working on a unit test for linalg.norm() (see ticket #1361), I discovered that if a vector is passed into the norm() method with 'fro' as the argument for the ord, an error occurs. Example: ``` >>> from numpy import linalg >>> a = [1,2,3,4] >>> linalg.norm(a,'fro') Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/tmp/lib/python2.5/site-packages/numpy/linalg/linalg.py", line 1262, in norm return ((abs(x)**ord).sum())**(1.0/ord) TypeError: unsupported operand type(s) for ** or pow(): 'numpy.ndarray' and 'str' >>> ``` Performing the same test but omiting the parameter 'fro' allows the method to execute normally. A test case that exposes this is available in ticket #1361.

2012-04-22T11:18:04Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/tmp/lib/python2.5/site-packages/numpy/linalg/linalg.py", line 1262, in norm return ((abs(x)**ord).sum())**(1.0/ord) TypeError: unsupported operand type(s) for ** or pow(): 'numpy.ndarray' and 'str'

10,219

numpy/numpy

numpy__numpy-3447

2a5c2c8227b600654f31ed346c73cce77bef554d

diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -5643,6 +5643,9 @@ def __iter__(self): else: yield d + def __len__(self): + return self._data.__len__() + def filled(self, fill_value=None): """ Return a copy with masked fields filled with a given value.

MaskedArray record: TypeError: len() of unsized object (Trac #2116) _Original ticket http://projects.scipy.org/numpy/ticket/2116 on 2012-04-26 by trac user mwtoews, assigned to @pierregm._ I'm getting a inconsistent error while trying to get the length of a record from a masked array, but only when the mask is enabled. See the example: ``` import numpy as np my_dtype = [('a','i'),('b','f')] mar = np.ma.zeros(5, my_dtype) # Length of the first record; no error raised assert len(mar[0]) == 2 # Change mask for one of the fields print(mar[0]) # (0, 0.0) mar.mask[0][0] = True print(mar[0]) # (--, 0.0) # Repeat same command as above to reveal this bug assert len(mar[0]) == 2 ``` Raises the error: ``` Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object ``` I am using !NumPy version 1.6.1 obtained from http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy and the version of Python is 2.5.1 [MSC v.1310 32 bit (Intel)].

2013-06-16T01:41:13Z

[]

Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object

10,263

numpy/numpy

numpy__numpy-3452

5ba9fead691439697c6d1c3768d4205cbe822bc2

diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -5593,6 +5593,9 @@ def __iter__(self): else: yield d + def __len__(self): + return self._data.__len__() + def filled(self, fill_value=None): """ Return a copy with masked fields filled with a given value. @@ -5930,9 +5933,10 @@ class _frommethod: Name of the method to transform. """ - def __init__(self, methodname): + def __init__(self, methodname, reversed=False): self.__name__ = methodname self.__doc__ = self.getdoc() + self.reversed = reversed # def getdoc(self): "Return the doc of the function (from the doc of the method)." @@ -5944,6 +5948,11 @@ def getdoc(self): return doc # def __call__(self, a, *args, **params): + if self.reversed: + args = list(args) + arr = args[0] + args[0] = a + a = arr # Get the method from the array (if possible) method_name = self.__name__ method = getattr(a, method_name, None) @@ -5960,7 +5969,7 @@ def __call__(self, a, *args, **params): all = _frommethod('all') anomalies = anom = _frommethod('anom') any = _frommethod('any') -compress = _frommethod('compress') +compress = _frommethod('compress', reversed=True) cumprod = _frommethod('cumprod') cumsum = _frommethod('cumsum') copy = _frommethod('copy')

MaskedArray record: TypeError: len() of unsized object (Trac #2116) _Original ticket http://projects.scipy.org/numpy/ticket/2116 on 2012-04-26 by trac user mwtoews, assigned to @pierregm._ I'm getting a inconsistent error while trying to get the length of a record from a masked array, but only when the mask is enabled. See the example: ``` import numpy as np my_dtype = [('a','i'),('b','f')] mar = np.ma.zeros(5, my_dtype) # Length of the first record; no error raised assert len(mar[0]) == 2 # Change mask for one of the fields print(mar[0]) # (0, 0.0) mar.mask[0][0] = True print(mar[0]) # (--, 0.0) # Repeat same command as above to reveal this bug assert len(mar[0]) == 2 ``` Raises the error: ``` Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object ``` I am using !NumPy version 1.6.1 obtained from http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy and the version of Python is 2.5.1 [MSC v.1310 32 bit (Intel)].

2013-06-17T01:40:29Z

[]

Traceback (most recent call last): File "<interactive input>", line 1, in <module> TypeError: len() of unsized object

10,267

numpy/numpy

numpy__numpy-3854

106306dc9b0e3aaf0543d78c7c20761ef03f0213

diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -368,10 +368,11 @@ def solve(a, b): gufunc = _umath_linalg.solve1 else: - if a.shape[-1] == 0 and b.shape[-2] == 0: - a = a.reshape(a.shape[:-1] + (1,)) - bc = broadcast(a, b) - return wrap(empty(bc.shape, dtype=result_t)) + if b.size == 0: + if (a.shape[-1] == 0 and b.shape[-2] == 0) or b.shape[-1] == 0: + a = a[:,:1].reshape(a.shape[:-1] + (1,)) + bc = broadcast(a, b) + return wrap(empty(bc.shape, dtype=result_t)) gufunc = _umath_linalg.solve

solve() fails on 0-sized axis With current 1.8, 5a0d09c: ``` import numpy as np np.linalg.solve(np.eye(3), np.zeros((3, 0))) ``` gives ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/njs/src/numpy/.tox/py27/local/lib/python2.7/site-packages/numpy/linalg/linalg.py", line 380, in solve r = gufunc(a, b, signature=signature, extobj=extobj) ValueError: cannot remove a zero-sized axis from an iterator ``` The result should be an array with shape `(3, 0)`. This is a regression from 1.7, hence marking as a blocker. Thanks to Jens Jørgen Mortensen for the original report: http://mail.scipy.org/pipermail/numpy-discussion/2013-October/067898.html

2013-10-02T16:46:10Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/njs/src/numpy/.tox/py27/local/lib/python2.7/site-packages/numpy/linalg/linalg.py", line 380, in solve r = gufunc(a, b, signature=signature, extobj=extobj) ValueError: cannot remove a zero-sized axis from an iterator

10,284

numpy/numpy

numpy__numpy-4276

bd5894b29b897f16da8a3d64e0df94e93d6b2d4a

diff --git a/numpy/core/_methods.py b/numpy/core/_methods.py --- a/numpy/core/_methods.py +++ b/numpy/core/_methods.py @@ -63,8 +63,10 @@ def _mean(a, axis=None, dtype=None, out=None, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -107,8 +109,10 @@ def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -118,7 +122,9 @@ def _std(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.sqrt(ret, out=ret) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(um.sqrt(ret)) + else: + ret = um.sqrt(ret) return ret

mean() (and median()) should work with "object" arrays With NumPy 1.8, `mean()` started to break when calculating the (global) mean of an array that contains objects (arrays with an object `dtype`). This also breaks `median()` on such arrays. Here is an example: ``` >>> numpy.arange(10).astype(object).mean() Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype' ``` Another example is case of numbers with uncertainties from the uncertainties package (https://github.com/lebigot/uncertainties/issues/22). I think that it would be better if NumPy did not assume that scalar results have a `dtype`, since arrays can contain objects that have a meaningful mean. I believe that such objects should not be forced to have a `dtype`, which is obviously NumPy specific (they even can't, for Python scalars like floats). Furthermore, a `dtype` is in principle not necessary for the calculation of the mean of such objects, so it would look strange if they had to have one. The problem is that `numpy.mean()` assumes that the intermediate result obtained has a `dtype` (with a `type` attribute). Therefore, I suggest that NumPy's `mean()` also handle arrays of objects that are not of the standard NumPy types (their `dtype` is object, and they contain objects that have a meaningful mean, like ints, floats, numbers with uncertainties, etc.).

introduced in f16b12e87 by @charris Hmmm, this is annoying. It seems to me like the only way to fix this is probably to see if `dtype` is given, and then use `np.dtype(dtype).type(...)` and otherwise just do the plain operation? Yeah, I was thinking along the same lines. I would be curious to see what the issue that prompted the change was, to see if I can come up with any kind of better suggestion (who knows): what was the problem, exactly? The scalar returns didn't preserve type, i.e., float32 would go to float64. That was on account of type precedence between scalars being different than type precedence between scalars and arrays. @charris do we care even about that? Or is it enough if the passed in dtype actually gets honored? I honestly have troubles to figure out a good method of preserving the type quite right for the scalar result. I now think we may have to just check for object dtype input (or passed in dtype). The most secure method I can think of would be a new keyword argument to the ufuncs to skip PyArray_Return (would probably be slower though), but unless that is useful elsewhere it is not worth the trouble either. Any progress on this one? It costed me a hour of debugging today. If doing this properly is hard, please consider fixing the error message so it is obvious what's wrong. @charris do you have time to have a look at this? I also think we accumulated enough fixes to warrant a 1.8.1 release if we add this and the C99 windows fix. thoughts? I'll get it done today sometime. Agree on 1.8.1, I came to that conclusion this morning. We should also fix the `divide` and `true_divide` ufuncs when the `dtype` is given. I need to think about this a bit more before putting up a fix.

2014-02-10T05:05:28Z

[]

Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype'

10,290

numpy/numpy

numpy__numpy-4297

95f7a469b1e9ce460e31c41e1bd897ceff396f6b

diff --git a/numpy/core/_methods.py b/numpy/core/_methods.py --- a/numpy/core/_methods.py +++ b/numpy/core/_methods.py @@ -63,8 +63,10 @@ def _mean(a, axis=None, dtype=None, out=None, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -107,8 +109,10 @@ def _var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.true_divide( ret, rcount, out=ret, casting='unsafe', subok=False) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(ret / rcount) + else: + ret = ret / rcount return ret @@ -118,7 +122,9 @@ def _std(a, axis=None, dtype=None, out=None, ddof=0, keepdims=False): if isinstance(ret, mu.ndarray): ret = um.sqrt(ret, out=ret) - else: + elif hasattr(ret, 'dtype'): ret = ret.dtype.type(um.sqrt(ret)) + else: + ret = um.sqrt(ret) return ret diff --git a/numpy/distutils/fcompiler/gnu.py b/numpy/distutils/fcompiler/gnu.py --- a/numpy/distutils/fcompiler/gnu.py +++ b/numpy/distutils/fcompiler/gnu.py @@ -35,13 +35,13 @@ class GnuFCompiler(FCompiler): def gnu_version_match(self, version_string): """Handle the different versions of GNU fortran compilers""" - m = re.match(r'GNU Fortran', version_string) + m = re.search(r'GNU Fortran', version_string) if not m: return None - m = re.match(r'GNU Fortran\s+95.*?([0-9-.]+)', version_string) + m = re.search(r'GNU Fortran\s+95.*?([0-9-.]+)', version_string) if m: return ('gfortran', m.group(1)) - m = re.match(r'GNU Fortran.*?\-?([0-9-.]+)', version_string) + m = re.search(r'GNU Fortran.*?\-?([0-9-.]+)', version_string) if m: v = m.group(1) if v.startswith('0') or v.startswith('2') or v.startswith('3'): diff --git a/numpy/f2py/__init__.py b/numpy/f2py/__init__.py --- a/numpy/f2py/__init__.py +++ b/numpy/f2py/__init__.py @@ -28,20 +28,20 @@ def compile(source, from numpy.distutils.exec_command import exec_command import tempfile if source_fn is None: - fname = os.path.join(tempfile.mktemp()+'.f') + f = tempfile.NamedTemporaryFile(suffix='.f') else: - fname = source_fn - - f = open(fname, 'w') - f.write(source) - f.close() - - args = ' -c -m %s %s %s'%(modulename, fname, extra_args) - c = '%s -c "import numpy.f2py as f2py2e;f2py2e.main()" %s' %(sys.executable, args) - s, o = exec_command(c) - if source_fn is None: - try: os.remove(fname) - except OSError: pass + f = open(source_fn, 'w') + + try: + f.write(source) + f.flush() + + args = ' -c -m %s %s %s'%(modulename, f.name, extra_args) + c = '%s -c "import numpy.f2py as f2py2e;f2py2e.main()" %s' % \ + (sys.executable, args) + s, o = exec_command(c) + finally: + f.close() return s from numpy.testing import Tester diff --git a/numpy/f2py/f2py2e.py b/numpy/f2py/f2py2e.py --- a/numpy/f2py/f2py2e.py +++ b/numpy/f2py/f2py2e.py @@ -91,7 +91,7 @@ --lower is assumed with -h key, and --no-lower without -h key. --build-dir <dirname> All f2py generated files are created in <dirname>. - Default is tempfile.mktemp(). + Default is tempfile.mkdtemp(). --overwrite-signature Overwrite existing signature file. @@ -428,7 +428,7 @@ def run_compile(): del sys.argv[i] else: remove_build_dir = 1 - build_dir = os.path.join(tempfile.mktemp()) + build_dir = tempfile.mkdtemp() _reg1 = re.compile(r'[-][-]link[-]') sysinfo_flags = [_m for _m in sys.argv[1:] if _reg1.match(_m)] diff --git a/numpy/lib/financial.py b/numpy/lib/financial.py --- a/numpy/lib/financial.py +++ b/numpy/lib/financial.py @@ -628,21 +628,29 @@ def irr(values): Examples -------- - >>> print round(np.irr([-100, 39, 59, 55, 20]), 5) + >>> round(irr([-100, 39, 59, 55, 20]), 5) 0.28095 + >>> round(irr([-100, 0, 0, 74]), 5) + -0.0955 + >>> round(irr([-100, 100, 0, -7]), 5) + -0.0833 + >>> round(irr([-100, 100, 0, 7]), 5) + 0.06206 + >>> round(irr([-5, 10.5, 1, -8, 1]), 5) + 0.0886 (Compare with the Example given for numpy.lib.financial.npv) """ res = np.roots(values[::-1]) - # Find the root(s) between 0 and 1 - mask = (res.imag == 0) & (res.real > 0) & (res.real <= 1) - res = res[mask].real + mask = (res.imag == 0) & (res.real > 0) if res.size == 0: return np.nan + res = res[mask].real + # NPV(rate) = 0 can have more than one solution so we return + # only the solution closest to zero. rate = 1.0/res - 1 - if rate.size == 1: - rate = rate.item() + rate = rate.item(np.argmin(np.abs(rate))) return rate def npv(rate, values): diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -18,11 +18,10 @@ DOCLINES = __doc__.split("\n") import os -import shutil import sys -import re import subprocess + if sys.version_info[:2] < (2, 6) or (3, 0) <= sys.version_info[0:2] < (3, 2): raise RuntimeError("Python version 2.6, 2.7 or >= 3.2 required.") @@ -31,6 +30,7 @@ else: import __builtin__ as builtins + CLASSIFIERS = """\ Development Status :: 5 - Production/Stable Intended Audience :: Science/Research @@ -47,24 +47,13 @@ Operating System :: MacOS """ -NAME = 'numpy' -MAINTAINER = "NumPy Developers" -MAINTAINER_EMAIL = "numpy-discussion@scipy.org" -DESCRIPTION = DOCLINES[0] -LONG_DESCRIPTION = "\n".join(DOCLINES[2:]) -URL = "http://www.numpy.org" -DOWNLOAD_URL = "http://sourceforge.net/projects/numpy/files/NumPy/" -LICENSE = 'BSD' -CLASSIFIERS = [_f for _f in CLASSIFIERS.split('\n') if _f] -AUTHOR = "Travis E. Oliphant et al." -AUTHOR_EMAIL = "oliphant@enthought.com" -PLATFORMS = ["Windows", "Linux", "Solaris", "Mac OS-X", "Unix"] MAJOR = 1 MINOR = 8 MICRO = 0 ISRELEASED = False VERSION = '%d.%d.%d' % (MAJOR, MINOR, MICRO) + # Return the git revision as a string def git_version(): def _minimal_ext_cmd(cmd): @@ -100,18 +89,7 @@ def _minimal_ext_cmd(cmd): builtins.__NUMPY_SETUP__ = True -def write_version_py(filename='numpy/version.py'): - cnt = """ -# THIS FILE IS GENERATED FROM NUMPY SETUP.PY -short_version = '%(version)s' -version = '%(version)s' -full_version = '%(full_version)s' -git_revision = '%(git_revision)s' -release = %(isrelease)s - -if not release: - version = full_version -""" +def get_version_info(): # Adding the git rev number needs to be done inside write_version_py(), # otherwise the import of numpy.version messes up the build under Python 3. FULLVERSION = VERSION @@ -131,6 +109,23 @@ def write_version_py(filename='numpy/version.py'): if not ISRELEASED: FULLVERSION += '.dev-' + GIT_REVISION[:7] + return FULLVERSION, GIT_REVISION + + +def write_version_py(filename='numpy/version.py'): + cnt = """ +# THIS FILE IS GENERATED FROM NUMPY SETUP.PY +short_version = '%(version)s' +version = '%(version)s' +full_version = '%(full_version)s' +git_revision = '%(git_revision)s' +release = %(isrelease)s + +if not release: + version = full_version +""" + FULLVERSION, GIT_REVISION = get_version_info() + a = open(filename, 'w') try: a.write(cnt % {'version': VERSION, @@ -140,6 +135,7 @@ def write_version_py(filename='numpy/version.py'): finally: a.close() + def configuration(parent_package='',top_path=None): from numpy.distutils.misc_util import Configuration @@ -155,8 +151,36 @@ def configuration(parent_package='',top_path=None): return config -def setup_package(): +def check_submodules(): + """ verify that the submodules are checked out and clean + use `git submodule update --init`; on failure + """ + if not os.path.exists('.git'): + return + with open('.gitmodules') as f: + for l in f: + if 'path' in l: + p = l.split('=')[-1].strip() + if not os.path.exists(p): + raise ValueError('Submodule %s missing' % p) + + + proc = subprocess.Popen(['git', 'submodule', 'status'], + stdout=subprocess.PIPE) + status, _ = proc.communicate() + status = status.decode("ascii", "replace") + for line in status.splitlines(): + if line.startswith('-') or line.startswith('+'): + raise ValueError('Submodule not clean: %s' % line) + +from distutils.command.sdist import sdist +class sdist_checked(sdist): + """ check submodules on sdist to prevent incomplete tarballs """ + def run(self): + check_submodules() + sdist.run(self) +def setup_package(): src_path = os.path.dirname(os.path.abspath(sys.argv[0])) old_path = os.getcwd() os.chdir(src_path) @@ -165,28 +189,51 @@ def setup_package(): # Rewrite the version file everytime write_version_py() + metadata = dict( + name = 'numpy', + maintainer = "NumPy Developers", + maintainer_email = "numpy-discussion@scipy.org", + description = DOCLINES[0], + long_description = "\n".join(DOCLINES[2:]), + url = "http://www.numpy.org", + author = "Travis E. Oliphant et al.", + download_url = "http://sourceforge.net/projects/numpy/files/NumPy/", + license = 'BSD', + classifiers=[_f for _f in CLASSIFIERS.split('\n') if _f], + platforms = ["Windows", "Linux", "Solaris", "Mac OS-X", "Unix"], + test_suite='nose.collector', + cmdclass={"sdist": sdist_checked}, + ) + # Run build - from numpy.distutils.core import setup + if len(sys.argv) >= 2 and ('--help' in sys.argv[1:] or + sys.argv[1] in ('--help-commands', 'egg_info', '--version', + 'clean')): + # Use setuptools for these commands (they don't work well or at all + # with distutils). For normal builds use distutils. + try: + from setuptools import setup + except ImportError: + from distutils.core import setup + + FULLVERSION, GIT_REVISION = get_version_info() + metadata['version'] = FULLVERSION + elif len(sys.argv) >= 2 and sys.argv[1] == 'bdist_wheel': + # bdist_wheel needs setuptools + import setuptools + from numpy.distutils.core import setup + metadata['configuration'] = configuration + else: + from numpy.distutils.core import setup + metadata['configuration'] = configuration try: - setup( - name=NAME, - maintainer=MAINTAINER, - maintainer_email=MAINTAINER_EMAIL, - description=DESCRIPTION, - long_description=LONG_DESCRIPTION, - url=URL, - download_url=DOWNLOAD_URL, - license=LICENSE, - classifiers=CLASSIFIERS, - author=AUTHOR, - author_email=AUTHOR_EMAIL, - platforms=PLATFORMS, - configuration=configuration ) + setup(**metadata) finally: del sys.path[0] os.chdir(old_path) return + if __name__ == '__main__': setup_package()

mean() (and median()) should work with "object" arrays With NumPy 1.8, `mean()` started to break when calculating the (global) mean of an array that contains objects (arrays with an object `dtype`). This also breaks `median()` on such arrays. Here is an example: ``` >>> numpy.arange(10).astype(object).mean() Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype' ``` Another example is case of numbers with uncertainties from the uncertainties package (https://github.com/lebigot/uncertainties/issues/22). I think that it would be better if NumPy did not assume that scalar results have a `dtype`, since arrays can contain objects that have a meaningful mean. I believe that such objects should not be forced to have a `dtype`, which is obviously NumPy specific (they even can't, for Python scalars like floats). Furthermore, a `dtype` is in principle not necessary for the calculation of the mean of such objects, so it would look strange if they had to have one. The problem is that `numpy.mean()` assumes that the intermediate result obtained has a `dtype` (with a `type` attribute). Therefore, I suggest that NumPy's `mean()` also handle arrays of objects that are not of the standard NumPy types (their `dtype` is object, and they contain objects that have a meaningful mean, like ints, floats, numbers with uncertainties, etc.).

introduced in f16b12e87 by @charris Hmmm, this is annoying. It seems to me like the only way to fix this is probably to see if `dtype` is given, and then use `np.dtype(dtype).type(...)` and otherwise just do the plain operation? Yeah, I was thinking along the same lines. I would be curious to see what the issue that prompted the change was, to see if I can come up with any kind of better suggestion (who knows): what was the problem, exactly? The scalar returns didn't preserve type, i.e., float32 would go to float64. That was on account of type precedence between scalars being different than type precedence between scalars and arrays. @charris do we care even about that? Or is it enough if the passed in dtype actually gets honored? I honestly have troubles to figure out a good method of preserving the type quite right for the scalar result. I now think we may have to just check for object dtype input (or passed in dtype). The most secure method I can think of would be a new keyword argument to the ufuncs to skip PyArray_Return (would probably be slower though), but unless that is useful elsewhere it is not worth the trouble either.

2014-02-15T17:25:25Z

[]

Traceback (most recent call last): File "<ipython-input-11-782b7c0104c3>", line 1, in <module> numpy.arange(10).astype(object).mean() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/numpy/core/_methods.py", line 67, in _mean ret = ret.dtype.type(ret / rcount) AttributeError: 'int' object has no attribute 'dtype'

10,291

numpy/numpy

numpy__numpy-4304

68ae04741f73953ef5680cb80fbb3dde5f160c90

diff --git a/numpy/lib/twodim_base.py b/numpy/lib/twodim_base.py --- a/numpy/lib/twodim_base.py +++ b/numpy/lib/twodim_base.py @@ -25,7 +25,7 @@ def fliplr(m): Parameters ---------- m : array_like - Input array. + Input array, must be at least 2-D. Returns ------- @@ -40,8 +40,7 @@ def fliplr(m): Notes ----- - Equivalent to A[:,::-1]. Does not require the array to be - two-dimensional. + Equivalent to A[:,::-1]. Requires the array to be at least 2-D. Examples --------

fliplr documentation incorrectly states 2-d *not* required [`fliplr` documentation](http://docs.scipy.org/doc/numpy/reference/generated/numpy.fliplr.html) incorrectly states 2-d _not_ required, but that is _not_ true. I think it is a typo, since `fliplr` was probably copied from `flipud` which _really_ doesn't require a 2-d array. ``` >>> import numpy as np >>> a = np.array([1,2,3,4]) >>> np.fliplr(a) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Python27\lib\site-packages\numpy\lib\twodim_base.py", line 61, in fliplr raise ValueError("Input must be >= 2-d.") ValueError: Input must be >= 2-d. ```

2014-02-16T20:32:54Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Python27\lib\site-packages\numpy\lib\twodim_base.py", line 61, in fliplr raise ValueError("Input must be >= 2-d.") ValueError: Input must be >= 2-d.

10,293

numpy/numpy

numpy__numpy-4305

68ae04741f73953ef5680cb80fbb3dde5f160c90

diff --git a/numpy/f2py/cfuncs.py b/numpy/f2py/cfuncs.py --- a/numpy/f2py/cfuncs.py +++ b/numpy/f2py/cfuncs.py @@ -312,7 +312,7 @@ needs['pyobj_from_string1']=['string'] cppmacros['pyobj_from_string1']='#define pyobj_from_string1(v) (PyString_FromString((char *)v))' needs['pyobj_from_string1size']=['string'] -cppmacros['pyobj_from_string1size']='#define pyobj_from_string1size(v,len) (PyString_FromStringAndSize((char *)v, len))' +cppmacros['pyobj_from_string1size']='#define pyobj_from_string1size(v,len) (PyUString_FromStringAndSize((char *)v, len))' needs['TRYPYARRAYTEMPLATE']=['PRINTPYOBJERR'] cppmacros['TRYPYARRAYTEMPLATE']="""\ /* New SciPy */

f2py sometimes generates python3-incompatible C wrappings f2py uses deprecated CPython functions (notably `PyString_FromStringAndSize`) when wrapping callbacks with string parameters. An example is the FORTRAN-77 subroutine ``` fortran SUBROUTINE BUG(CALLBACK, A) IMPLICIT NONE EXTERNAL CALLBACK DOUBLE PRECISION CALLBACK DOUBLE PRECISION A CHARACTER*1 R R = 'R' !f2py intent(out) A A = CALLBACK(R) END SUBROUTINE ``` compiled with ``` f2py -c -m bug test.f # on Arch Linux, this is the python3 version of f2py ``` with the Python wrapping ``` python #!/usr/bin/env python from __future__ import print_function import bug def callback(code): if code == 'F': return 2.0 else: return 3.0 print(bug.bug(callback)) ``` Calling the Python2 version of `f2py` and running the script under Python 2.7 yields the correct answer. However, under Python3 we get ``` ~/Documents/Code/Fortran/f2pyBug $ python3 test.py Traceback (most recent call last): File "test.py", line 4, in <module> import bug ImportError: /home/drwells/Documents/Code/Fortran/f2pyBug/bug.cpython-33m.so: undefined symbol: PyString_FromStringAndSize ``` I tested this under Arch Linux (fully updated) with Python 3.3.3 and Python 2.7.6. In both cases I used Numpy 1.8. I also tested this under Ubuntu (Quantal) with Python 3.2.3 and Python 2.7.3. In both cases I used Numpy 1.6.2.

Fyi, unfortunately f2py is not regularly maintained at this point, so you may or may not find someone stepping up to fix it. Your best bet might be to see if you can figure out a patch yourself - we'd definitely be interested in merging such a thing. On 3 Feb 2014 13:51, "drwells" notifications@github.com wrote: > f2py uses deprecated CPython functions (notably PyString_FromStringAndSize) > when wrapping callbacks with string parameters. An example is the > FORTRAN-77 subroutine > > ``` > SUBROUTINE BUG(CALLBACK, A) > IMPLICIT NONE EXTERNAL CALLBACK > DOUBLE PRECISION CALLBACK > DOUBLE PRECISION A > CHARACTER*1 R > R = 'R'!f2py intent(out) A > A = CALLBACK(R) > END SUBROUTINE > ``` > > compiled with > > f2py -c -m bug test.f # on Arch Linux, this is the python3 version of f2py > > with the Python wrapping > > #!/usr/bin/env pythonfrom **future** import print_functionimport bug > def callback(code): > if code == 'F': > return 2.0 > else: > return 3.0 > print(bug.bug(callback)) > > Calling the Python2 version of f2py and running the script under Python > 2.7 yields the correct answer. However, under Python3 we get > > ~/Documents/Code/Fortran/f2pyBug $ python3 test.py > Traceback (most recent call last): > File "test.py", line 4, in <module> > import bug > ImportError: /home/drwells/Documents/Code/Fortran/f2pyBug/bug.cpython-33m.so: undefined symbol: PyString_FromStringAndSize > > I tested this under Arch Linux (fully updated) with Python 3.3.3 and > Python 2.7.6. In both cases I used Numpy 1.8. I also tested this under > Ubuntu (Quantal) with Python 3.2.3 and Python 2.7.3. In both cases I used > Numpy 1.6.2. > > — > Reply to this email directly or view it on GitHubhttps://github.com/numpy/numpy/issues/4256 > . These are likely easy to fix by adding the missing defines to numpy/f2py/src/fortranobject.h

2014-02-16T22:17:50Z

[]

Traceback (most recent call last): File "test.py", line 4, in <module> import bug ImportError: /home/drwells/Documents/Code/Fortran/f2pyBug/bug.cpython-33m.so: undefined symbol: PyString_FromStringAndSize

10,294

numpy/numpy

numpy__numpy-4479

42be275997e56b7e21d3adab5d5a5142876db9f0

diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -1593,6 +1593,7 @@ def __init__(self, pyfunc, otypes='', doc=None, excluded=None, cache=False): self.pyfunc = pyfunc self.cache = cache + self._ufunc = None # Caching to improve default performance if doc is None: self.__doc__ = pyfunc.__doc__ @@ -1616,9 +1617,6 @@ def __init__(self, pyfunc, otypes='', doc=None, excluded=None, excluded = set() self.excluded = set(excluded) - if self.otypes and not self.excluded: - self._ufunc = None # Caching to improve default performance - def __call__(self, *args, **kwargs): """ Return arrays with the results of `pyfunc` broadcast (vectorized) over @@ -1652,7 +1650,8 @@ def func(*vargs): def _get_ufunc_and_otypes(self, func, args): """Return (ufunc, otypes).""" # frompyfunc will fail if args is empty - assert args + if not args: + raise ValueError('args can not be empty') if self.otypes: otypes = self.otypes

_vectorize_call error when otype attribute is set after vectorize object creation if a vectorize object is created without otype argument, an attempt to set it later causes an error: ``` python v = numpy.vectorize( lambda x: x ) v.otypes='i' v( [1,2] ) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1873, in __call__ return self._vectorize_call(func=func, args=vargs) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1933, in _vectorize_call ufunc, otypes = self._get_ufunc_and_otypes(func=func, args=args) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1886, in _get_ufunc_and_otypes if func is self.pyfunc and self._ufunc is not None: AttributeError: 'vectorize' object has no attribute '_ufunc' ```

Still in 1.9-devel. `_get_ufunc_and_otypes` is buggy in other ways, it uses `assert` for flow control.

2014-03-11T08:20:58Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1873, in __call__ return self._vectorize_call(func=func, args=vargs) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1933, in _vectorize_call ufunc, otypes = self._get_ufunc_and_otypes(func=func, args=args) File "/usr/lib64/python2.7/site-packages/numpy/lib/function_base.py", line 1886, in _get_ufunc_and_otypes if func is self.pyfunc and self._ufunc is not None: AttributeError: 'vectorize' object has no attribute '_ufunc'

10,310

numpy/numpy

numpy__numpy-4588

a0794f63d548e688e2eed76a9dc4e8df0ea33846

diff --git a/numpy/lib/npyio.py b/numpy/lib/npyio.py --- a/numpy/lib/npyio.py +++ b/numpy/lib/npyio.py @@ -845,6 +845,11 @@ def split_line(line): continue if usecols: vals = [vals[i] for i in usecols] + if len(vals) != N: + line_num = i + skiprows + 1 + raise ValueError("Wrong number of columns at line %d" + % line_num) + # Convert each value according to its column and store items = [conv(val) for (conv, val) in zip(converters, vals)] # Then pack it according to the dtype's nesting

loadtxt should give line numbers of problems (Trac #1998) _Original ticket http://projects.scipy.org/numpy/ticket/1998 on 2011-12-15 by @samtygier, assigned to unknown._ currently input like ``` 1 2 3 4 5 6 78 ``` will cause loadtxt to give the following error ``` Traceback (most recent call last): File "./np_lt.py", line 5, in <module> numpy.loadtxt(sys.argv[1]) File "/usr/lib/python2.7/site-packages/numpy/lib/npyio.py", line 804, in loadtxt X = np.array(X, dtype) ValueError: setting an array element with a sequence. ``` This would be far more useful, if it contained some information about where the problem was. This is similar to the bug #1810 , but for the loadtxt function. I attach a patch that gives the following message: ``` ValueError: Wrong number of columns at line 3 ``` it also provides a test that the input raises the correct error. Inconveniently it is the same error type as before, so the test is not very useful. However I thing ValueError is probably the most appropriate in this case. Maybe there is a way to check the exception message.

Attachment added by @samtygier on 2011-12-15: [numpy_loadtxt.diff](http://projects.scipy.org/numpy/attachment/ticket/1998/numpy_loadtxt.diff) Good idea.

2014-04-05T05:03:53Z

[]

Traceback (most recent call last): File "./np_lt.py", line 5, in <module> numpy.loadtxt(sys.argv[1]) File "/usr/lib/python2.7/site-packages/numpy/lib/npyio.py", line 804, in loadtxt X = np.array(X, dtype) ValueError: setting an array element with a sequence.

10,314

numpy/numpy

numpy__numpy-4677

8e0ac440329188b959520a0a7ce41ef60b2fb3c2

diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -7018,8 +7018,6 @@ def asarray(a, dtype=None, order=None): <class 'numpy.ma.core.MaskedArray'> """ - if dtype is None and type(a) is MaskedArray: - return a return masked_array(a, dtype=dtype, copy=False, keep_mask=True, subok=False) def asanyarray(a, dtype=None): @@ -7065,8 +7063,6 @@ def asanyarray(a, dtype=None): <class 'numpy.ma.core.MaskedArray'> """ - if dtype is None and isinstance(a, MaskedArray): - return a return masked_array(a, dtype=dtype, copy=False, keep_mask=True, subok=True)

Test error for scipy 0.15.0.dev-5d197ed The first bad commit is d8fd28389adb491e24b7cdc25cd1b20f539310c3, the isolated error is ``` Traceback (most recent call last): File "/home/charris/scipy-test-fail.py", line 53, in test_trim assert_equal(trimx._mask.ravel(), expected) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 123, in assert_equal return assert_array_equal(actual, desired, err_msg) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 196, in assert_array_equal header='Arrays are not equal') File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 189, in assert_array_compare verbose=verbose, header=header) File "/home/charris/.local/lib/python2.7/site-packages/numpy/testing/utils.py", line 651, in assert_array_compare raise AssertionError(msg) AssertionError: Arrays are not equal (mismatch 9.09090909091%) x: array([ True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, True, False, False, False, False, False, False, False,... y: array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,... ---------------------------------------------------------------------- Ran 1 test in 0.005s ```

Looks like the problem is returning a view rather than a copy. Or more precisely, not making a copy and returning the original array. Probably the easiest fix is to revert this change, at least for numpy 1.9.

2014-05-05T17:28:43Z

[]

Traceback (most recent call last): File "/home/charris/scipy-test-fail.py", line 53, in test_trim assert_equal(trimx._mask.ravel(), expected) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 123, in assert_equal return assert_array_equal(actual, desired, err_msg) File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 196, in assert_array_equal header='Arrays are not equal') File "/home/charris/.local/lib/python2.7/site-packages/numpy/ma/testutils.py", line 189, in assert_array_compare verbose=verbose, header=header) File "/home/charris/.local/lib/python2.7/site-packages/numpy/testing/utils.py", line 651, in assert_array_compare raise AssertionError(msg) AssertionError:

10,318

numpy/numpy

numpy__numpy-4792

db710cefeecf51d6253e421712726c1506a6f65b

diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -651,7 +651,7 @@ def piecewise(x, condlist, funclist, *args, **kw): The output is the same shape and type as x and is found by calling the functions in `funclist` on the appropriate portions of `x`, as defined by the boolean arrays in `condlist`. Portions not covered - by any condition have undefined values. + by any condition have a default value of 0. See Also @@ -693,32 +693,24 @@ def piecewise(x, condlist, funclist, *args, **kw): if (isscalar(condlist) or not (isinstance(condlist[0], list) or isinstance(condlist[0], ndarray))): condlist = [condlist] - condlist = [asarray(c, dtype=bool) for c in condlist] + condlist = array(condlist, dtype=bool) n = len(condlist) - if n == n2 - 1: # compute the "otherwise" condition. - totlist = condlist[0] - for k in range(1, n): - totlist |= condlist[k] - condlist.append(~totlist) - n += 1 - if (n != n2): - raise ValueError( - "function list and condition list must be the same") - zerod = False # This is a hack to work around problems with NumPy's # handling of 0-d arrays and boolean indexing with # numpy.bool_ scalars + zerod = False if x.ndim == 0: x = x[None] zerod = True - newcondlist = [] - for k in range(n): - if condlist[k].ndim == 0: - condition = condlist[k][None] - else: - condition = condlist[k] - newcondlist.append(condition) - condlist = newcondlist + if condlist.shape[-1] != 1: + condlist = condlist.T + if n == n2 - 1: # compute the "otherwise" condition. + totlist = np.logical_or.reduce(condlist, axis=0) + condlist = np.vstack([condlist, ~totlist]) + n += 1 + if (n != n2): + raise ValueError( + "function list and condition list must be the same") y = zeros(x.shape, x.dtype) for k in range(n):

Fixed bug in numpy.piecewise() for 0-d array handling Updated numpy/lib/function_base.py to fix bug in numpy.piecewise() for 0-d array handling and boolean indexing with scalars. Bug test case: ``` >>> numpy.piecewise(5, [True, False], [1, 0]) Traceback (most recent call last): [...] y[condlist[k]] = item ValueError: boolean index array has too many values ``` After fix: ``` >>> numpy.piecewise(5, [True, False], [1, 0]) array(1) ```

This pull request [fails](http://travis-ci.org/numpy/numpy/builds/1788960) (merged e553e1b4 into 731cf3aa). The current version of this pull request appears to fail the existing tests (e.g., this build reports failures: http://travis-ci.org/#!/numpy/numpy/jobs/1788964), so that would need to be fixed. Also, some sort of test needs to be added to make sure that the bug is fixed (and stays that way). Hi njsmith, thanks. I've fixed both my change to numpy.piecewise() as well as the test cases to cover both my bug case above as well as making sure that the tests are all logical. This pull request [passes](http://travis-ci.org/numpy/numpy/builds/1804356) (merged e2ad1a72 into 731cf3aa). I guess you rearranged the order of all the tests in the source? It makes the diff unreadable unfortunately -- I can't tell at all what you've actually changed without reverse-engineering it by hand, and I'm afraid I'm too lazy for that. Can you revert the no-op changes, or split them into a separate commit, or at least describe what changes were actually made? Is this issue (i.e. this pull request) going to be fixed anytime soon? It's a bit anoying to construct a 1d array with one element each time I want to call this function with a scalar. I guess it will be fixed as soon as someone does the work :-) Hm, got it :) Should an issue be filed first? The main point of filing an issue is to make sure that something doesn't get forgotten, and this PR is already sort of doing that (i.e., it shows up on the list of open issues if anyone goes to look at it). But feel free to file one if you like. If you're thinking about fixing the bug yourself, then we're pretty laid back about these things -- you can just go ahead and submit a new PR without filing an issue if you want. We'd rather get the fix than worry about the process :-). The main thing that needs done is just to rewrite the unit tests. In my own testing and usage (and when I've tested with the original unit tests), the code works fine, but I just haven't had the time to sit down and rewrite them to be more similar to the original set and still be complete and systematic. I may get to it next week sometime, but until then I'm swamped. -- Eric On Nov 15, 2012, at 9:33 AM, njsmith notifications@github.com wrote: > The main point of filing an issue is to make sure that something doesn't get forgotten, and this PR is already sort of doing that (i.e., it shows up on the list of open issues if anyone goes to look at it). But feel free to file one if you like. If you're thinking about fixing the bug yourself, then we're pretty laid back about these things -- you can just go ahead and submit a new PR without filing an issue if you want. We'd rather get the fix than worry about the process :-). > > — > Reply to this email directly or view it on GitHub. That's one of the things I was about to ask, it is ok to rewrite the tests? Because right now IMHO they don't seem very meaningful for the purpose of the function. I'd like to try it myself this weekend, or next one as much. It's OK to make any changes whatsoever, the requirements are just that you should have a reason, and we should be able to look at your changes and figure out if they're good :-). How exactly to do that depends on the change. Easily readable diffs, English explanations of what you did and why... whatever works. On 17 Nov 2012 00:02, "Juan Luis Cano Rodríguez" notifications@github.com wrote: > That's one of the things I was about to ask, it is ok to rewrite the > tests? Because right now IMHO they don't seem very meaningful for the > purpose of the function. I'd like to try it myself this weekend, or next > one as much. > > — > Reply to this email directly or view it on GitHubhttps://github.com/numpy/numpy/pull/331#issuecomment-10467386. What is the status of this? Well actually I didn't do it any of the weekends as you can notice :( But anyway now that you ask about it, if I recall correctly there was some discussion recently about rank-0 arrays in the mailing list but I am not able to find it. Are there any constraints on how to do this? Other than that it would be my first contribution to NumPy and I am a bit unsecure but I am willing to give it a try. Yes, I was referring to this: http://www.mail-archive.com/numpy-discussion@scipy.org/msg39968.html My question was if there are any near future plans for rank-0 arrays that relate this and other issues in such a way that it requires changing low-level code or if we can address this individually. No, I don't think there will be any near-future changes in how 0d arrays work that will affect this bug. On 1 Mar 2013 07:29, "Juan Luis Cano Rodríguez" notifications@github.com wrote: > Yes, I was referring to this: > > http://www.mail-archive.com/numpy-discussion@scipy.org/msg39968.html > > My question was if there are any near future plans for rank-0 arrays that > relate this and other issues in such a way that it requires changing > low-level code or if we can address this individually. > > — > Reply to this email directly or view it on GitHubhttps://github.com/numpy/numpy/pull/331#issuecomment-14276274 > . Sorry for the delay of months. Finishing up my degree and job hunting, so I've been...occupied. I reverted most of the unit test changes so that it's clearer what I changed. Maybe I'll re-up the unit test changes at some point, but this particular commit should be easier to understand when you diff between this state and the state 3 commits ago. @njsmith Can this go in. Ping Travisbot just because it was a long time ago. @ericsuh Made some comments. The commits should also be squashed into one. I've put up a cleaned version at https://github.com/charris/numpy in the branch gh-331 Needs finishing up. I seem to have deleted my cleanup somewhere along the line. In the docstring: "Each boolean array in `condlist` selects a piece of `x`, and should therefore be of the same shape as `x`." But actually even the tests do not respect this (and still everything works): https://github.com/numpy/numpy/blob/b1c69df01b673cc086065112da6780d8548a0dfa/numpy/lib/tests/test_function_base.py#L1468 There are other bugs too. For instance, there's `test_0d` but even though `np.piecewise(x, x > 3, [4, 0])` passes, `np.piecewise(x, [x > 3, x <= 3], [4, 0])` fails (same error as originally reported). I don't think this behaviour is consistent - that's what I actually meant a year and a half ago. The precondition stated in the docstring makes sense for me and should be somehow enforced - is anybody against it? I cannot promise anything but I may try and solve this at once.

2014-06-07T20:44:26Z

[]

Traceback (most recent call last): [...] y[condlist[k]] = item ValueError: boolean index array has too many values

10,321

numpy/numpy

numpy__numpy-5149

58350f4608a22f4b4b66795f51eaefc206bd02b8

diff --git a/numpy/ma/extras.py b/numpy/ma/extras.py --- a/numpy/ma/extras.py +++ b/numpy/ma/extras.py @@ -434,8 +434,10 @@ def apply_over_axes(func, a, axes): raise ValueError("function is not returning " "an array of the correct shape") return val -apply_over_axes.__doc__ = np.apply_over_axes.__doc__[ - :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ + +if apply_over_axes.__doc__ is not None: + apply_over_axes.__doc__ = np.apply_over_axes.__doc__[ + :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ """ Examples @@ -462,7 +464,7 @@ def apply_over_axes(func, a, axes): [[[46] [--] [124]]] -""" + """ def average(a, axis=None, weights=None, returned=False):

Numpy crashes with -OO Hello, I ran `$ python -OO -c 'import numpy'` on a fresh miniconda install (after `$ conda install numpy`) and I got the following output: ``` Traceback (most recent call last): File "<string>", line 1, in <module> File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/__init__.py", line 191, in <module> from . import ma File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/__init__.py", line 49, in <module> from . import extras File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/extras.py", line 438, in <module> :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ AttributeError: 'NoneType' object has no attribute 'find' ``` Running `$ python` gives me the following output (I am running OS X Mavericks on a Dec-2010 MacBook Pro): ``` Python 2.7.8 |Continuum Analytics, Inc.| (default, Aug 21 2014, 15:21:46) [GCC 4.2.1 (Apple Inc. build 5577)] on darwin Type "help", "copyright", "credits" or "license" for more information. Anaconda is brought to you by Continuum Analytics. Please check out: http://continuum.io/thanks and https://binstar.org ``` Finally, running `python -c 'import numpy'` works normally. I hope this is enough information for you all, but please let me know if you need more. (It's the second bug report I submit in my life, and the first one was about a typo in documentation!) Thanks a lot for making numpy available to us, it is a very, very useful tool that works greatly for me.

2014-10-04T02:12:20Z

[]

Traceback (most recent call last): File "<string>", line 1, in <module> File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/__init__.py", line 191, in <module> from . import ma File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/__init__.py", line 49, in <module> from . import extras File "/Users/obernardo/miniconda/lib/python2.7/site-packages/numpy/ma/extras.py", line 438, in <module> :np.apply_over_axes.__doc__.find('Notes')].rstrip() + \ AttributeError: 'NoneType' object has no attribute 'find'

10,330

numpy/numpy

numpy__numpy-5455

14445500bdf67600f926c6426bad55977441dca0

diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -145,10 +145,15 @@ class MaskError(MAError): 'S' : 'N/A', 'u' : 999999, 'V' : '???', - 'U' : 'N/A', - 'M8[D]' : np.datetime64('NaT', 'D'), - 'M8[us]' : np.datetime64('NaT', 'us') + 'U' : 'N/A' } + +# Add datetime64 and timedelta64 types +for v in ["Y", "M", "W", "D", "h", "m", "s", "ms", "us", "ns", "ps", + "fs", "as"]: + default_filler["M8[" + v + "]"] = np.datetime64("NaT", v) + default_filler["m8[" + v + "]"] = np.timedelta64("NaT", v) + max_filler = ntypes._minvals max_filler.update([(k, -np.inf) for k in [np.float32, np.float64]]) min_filler = ntypes._maxvals @@ -194,7 +199,7 @@ def default_fill_value(obj): 999999 >>> np.ma.default_fill_value(np.array([1.1, 2., np.pi])) 1e+20 - >>> np.ma.default_fill_value(np.dtype(complex)) + >>> np.ma.default_fill_value(np.dtype(complex)) (1e+20+0j) """ @@ -203,7 +208,7 @@ def default_fill_value(obj): elif isinstance(obj, np.dtype): if obj.subdtype: defval = default_filler.get(obj.subdtype[0].kind, '?') - elif obj.kind == 'M': + elif obj.kind in 'Mm': defval = default_filler.get(obj.str[1:], '?') else: defval = default_filler.get(obj.kind, '?')

Masked array view fails if structured dtype has datetime component A view as `numpy.ma.MaskedArray` fails if the array has a structured dtype, including at least one part that is `datetime64`, as follows: ``` $ python3.3 Python 3.3.2+ (default, Feb 28 2014, 00:52:16) [GCC 4.8.1] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import numpy >>> import numpy, numpy.ma, numpy.version >>> numpy.version.version '1.8.0' >>> A = numpy.empty(shape=(5,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")]) >>> A.view(numpy.ma.MaskedArray) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0 ```

A simpler `numpy.ma.MaskedArray([], shape=(0,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")])` fails with the same `ValueError`. Just tested with with the latest git repository. Problem still exists.

2015-01-15T00:07:57Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0

10,346

numpy/numpy

numpy__numpy-5498

54ebec335384ec1d6f8e65bdd35c4f359797dc0b

diff --git a/numpy/add_newdocs.py b/numpy/add_newdocs.py --- a/numpy/add_newdocs.py +++ b/numpy/add_newdocs.py @@ -885,7 +885,7 @@ def luf(lamdaexpr, *args, **kwargs): >>> np.zeros(5) array([ 0., 0., 0., 0., 0.]) - >>> np.zeros((5,), dtype=numpy.int) + >>> np.zeros((5,), dtype=np.int) array([0, 0, 0, 0, 0]) >>> np.zeros((2, 1))

Error in help documentation of numpy.zeros The help documentation for numpy zeros method (`import numpy; help(numpy.zeros)`) shows an example like this: ``` python >>> np.zeros((5,), dtype=numpy.int) array([0, 0, 0, 0, 0]) ``` The documentation throughout `multiarray` module assumes that `numpy` is imported as `np` but the value for `dtype` keyword is passed as `numpy.int`. While the above snippet conveys the point across, it will throw a name error if some one actually tried to execute it: ``` python >>> import numpy as np >>> np.zeros((5,), dtype=numpy.int) Traceback (most recent call last): File "<stdin>", line 1, in <module> NameError: name 'numpy' is not defined ``` This can be fixed by consitent usage of `numpy` or `np`.

2015-01-24T09:50:00Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> NameError: name 'numpy' is not defined

10,351

numpy/numpy

numpy__numpy-5500

1cac77b27d10217aaeb468abb71ed49f241ce469

diff --git a/numpy/lib/npyio.py b/numpy/lib/npyio.py --- a/numpy/lib/npyio.py +++ b/numpy/lib/npyio.py @@ -717,7 +717,8 @@ def loadtxt(fname, dtype=float, comments='#', delimiter=None, """ # Type conversions for Py3 convenience - comments = asbytes(comments) + if comments is not None: + comments = asbytes(comments) user_converters = converters if delimiter is not None: delimiter = asbytes(delimiter) @@ -790,7 +791,10 @@ def pack_items(items, packing): def split_line(line): """Chop off comments, strip, and split at delimiter.""" - line = asbytes(line).split(comments)[0].strip(asbytes('\r\n')) + if comments is None: + line = asbytes(line).strip(asbytes('\r\n')) + else: + line = asbytes(line).split(comments)[0].strip(asbytes('\r\n')) if line: return line.split(delimiter) else: diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -145,10 +145,15 @@ class MaskError(MAError): 'S' : 'N/A', 'u' : 999999, 'V' : '???', - 'U' : 'N/A', - 'M8[D]' : np.datetime64('NaT', 'D'), - 'M8[us]' : np.datetime64('NaT', 'us') + 'U' : 'N/A' } + +# Add datetime64 and timedelta64 types +for v in ["Y", "M", "W", "D", "h", "m", "s", "ms", "us", "ns", "ps", + "fs", "as"]: + default_filler["M8[" + v + "]"] = np.datetime64("NaT", v) + default_filler["m8[" + v + "]"] = np.timedelta64("NaT", v) + max_filler = ntypes._minvals max_filler.update([(k, -np.inf) for k in [np.float32, np.float64]]) min_filler = ntypes._maxvals @@ -194,7 +199,7 @@ def default_fill_value(obj): 999999 >>> np.ma.default_fill_value(np.array([1.1, 2., np.pi])) 1e+20 - >>> np.ma.default_fill_value(np.dtype(complex)) + >>> np.ma.default_fill_value(np.dtype(complex)) (1e+20+0j) """ @@ -203,7 +208,7 @@ def default_fill_value(obj): elif isinstance(obj, np.dtype): if obj.subdtype: defval = default_filler.get(obj.subdtype[0].kind, '?') - elif obj.kind == 'M': + elif obj.kind in 'Mm': defval = default_filler.get(obj.str[1:], '?') else: defval = default_filler.get(obj.kind, '?')

loadtxt(comments=None) considers the string 'None' as a comment symbol Numpy 1.8.2 `numpy.loadtxt(comments=None)` considers the string `'None'` as a comment symbol. Expected behaviour: `comments=None` should indicate that there are no comments symbols. Certainly not cast `None` to a string, what is this, Javascript? Masked array view fails if structured dtype has datetime component A view as `numpy.ma.MaskedArray` fails if the array has a structured dtype, including at least one part that is `datetime64`, as follows: ``` $ python3.3 Python 3.3.2+ (default, Feb 28 2014, 00:52:16) [GCC 4.8.1] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import numpy >>> import numpy, numpy.ma, numpy.version >>> numpy.version.version '1.8.0' >>> A = numpy.empty(shape=(5,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")]) >>> A.view(numpy.ma.MaskedArray) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0 ```

A simpler `numpy.ma.MaskedArray([], shape=(0,), dtype=[("A", "<f4"), ("B", "datetime64[ms]")])` fails with the same `ValueError`. Just tested with with the latest git repository. Problem still exists.

2015-01-25T16:09:27Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 2800, in __array_finalize__ self._fill_value = _check_fill_value(None, self.dtype) File "/home/gerrit/venv/python-3.3/lib/python3.3/site-packages/numpy/ma/core.py", line 402, in _check_fill_value dtype=ndtype,) ValueError: Error parsing datetime string "?" at position 0

10,352

numpy/numpy

numpy__numpy-5548

30d755d8737505717d54ed32501261bb94130a7f

diff --git a/numpy/core/_internal.py b/numpy/core/_internal.py --- a/numpy/core/_internal.py +++ b/numpy/core/_internal.py @@ -305,6 +305,174 @@ def _index_fields(ary, fields): copy_dtype = {'names':view_dtype['names'], 'formats':view_dtype['formats']} return array(view, dtype=copy_dtype, copy=True) +def _get_all_field_offsets(dtype, base_offset=0): + """ Returns the types and offsets of all fields in a (possibly structured) + data type, including nested fields and subarrays. + + Parameters + ---------- + dtype : data-type + Data type to extract fields from. + base_offset : int, optional + Additional offset to add to all field offsets. + + Returns + ------- + fields : list of (data-type, int) pairs + A flat list of (dtype, byte offset) pairs. + + """ + fields = [] + if dtype.fields is not None: + for name in dtype.names: + sub_dtype = dtype.fields[name][0] + sub_offset = dtype.fields[name][1] + base_offset + fields.extend(_get_all_field_offsets(sub_dtype, sub_offset)) + else: + if dtype.shape: + sub_offsets = _get_all_field_offsets(dtype.base, base_offset) + count = 1 + for dim in dtype.shape: + count *= dim + fields.extend((typ, off + dtype.base.itemsize*j) + for j in range(count) for (typ, off) in sub_offsets) + else: + fields.append((dtype, base_offset)) + return fields + +def _check_field_overlap(new_fields, old_fields): + """ Perform object memory overlap tests for two data-types (see + _view_is_safe). + + This function checks that new fields only access memory contained in old + fields, and that non-object fields are not interpreted as objects and vice + versa. + + Parameters + ---------- + new_fields : list of (data-type, int) pairs + Flat list of (dtype, byte offset) pairs for the new data type, as + returned by _get_all_field_offsets. + old_fields: list of (data-type, int) pairs + Flat list of (dtype, byte offset) pairs for the old data type, as + returned by _get_all_field_offsets. + + Raises + ------ + TypeError + If the new fields are incompatible with the old fields + + """ + from .numerictypes import object_ + from .multiarray import dtype + + #first go byte by byte and check we do not access bytes not in old_fields + new_bytes = set() + for tp, off in new_fields: + new_bytes.update(set(range(off, off+tp.itemsize))) + old_bytes = set() + for tp, off in old_fields: + old_bytes.update(set(range(off, off+tp.itemsize))) + if new_bytes.difference(old_bytes): + raise TypeError("view would access data parent array doesn't own") + + #next check that we do not interpret non-Objects as Objects, and vv + obj_offsets = [off for (tp, off) in old_fields if tp.type is object_] + obj_size = dtype(object_).itemsize + + for fld_dtype, fld_offset in new_fields: + if fld_dtype.type is object_: + # check we do not create object views where + # there are no objects. + if fld_offset not in obj_offsets: + raise TypeError("cannot view non-Object data as Object type") + else: + # next check we do not create non-object views + # where there are already objects. + # see validate_object_field_overlap for a similar computation. + for obj_offset in obj_offsets: + if (fld_offset < obj_offset + obj_size and + obj_offset < fld_offset + fld_dtype.itemsize): + raise TypeError("cannot view Object as non-Object type") + +def _getfield_is_safe(oldtype, newtype, offset): + """ Checks safety of getfield for object arrays. + + As in _view_is_safe, we need to check that memory containing objects is not + reinterpreted as a non-object datatype and vice versa. + + Parameters + ---------- + oldtype : data-type + Data type of the original ndarray. + newtype : data-type + Data type of the field being accessed by ndarray.getfield + offset : int + Offset of the field being accessed by ndarray.getfield + + Raises + ------ + TypeError + If the field access is invalid + + """ + new_fields = _get_all_field_offsets(newtype, offset) + old_fields = _get_all_field_offsets(oldtype) + # raises if there is a problem + _check_field_overlap(new_fields, old_fields) + +def _view_is_safe(oldtype, newtype): + """ Checks safety of a view involving object arrays, for example when + doing:: + + np.zeros(10, dtype=oldtype).view(newtype) + + We need to check that + 1) No memory that is not an object will be interpreted as a object, + 2) No memory containing an object will be interpreted as an arbitrary type. + Both cases can cause segfaults, eg in the case the view is written to. + Strategy here is to also disallow views where newtype has any field in a + place oldtype doesn't. + + Parameters + ---------- + oldtype : data-type + Data type of original ndarray + newtype : data-type + Data type of the view + + Raises + ------ + TypeError + If the new type is incompatible with the old type. + + """ + new_fields = _get_all_field_offsets(newtype) + new_size = newtype.itemsize + + old_fields = _get_all_field_offsets(oldtype) + old_size = oldtype.itemsize + + # if the itemsizes are not equal, we need to check that all the + # 'tiled positions' of the object match up. Here, we allow + # for arbirary itemsizes (even those possibly disallowed + # due to stride/data length issues). + if old_size == new_size: + new_num = old_num = 1 + else: + gcd_new_old = _gcd(new_size, old_size) + new_num = old_size // gcd_new_old + old_num = new_size // gcd_new_old + + # get position of fields within the tiling + new_fieldtile = [(tp, off + new_size*j) + for j in range(new_num) for (tp, off) in new_fields] + old_fieldtile = [(tp, off + old_size*j) + for j in range(old_num) for (tp, off) in old_fields] + + # raises if there is a problem + _check_field_overlap(new_fieldtile, old_fieldtile) + # Given a string containing a PEP 3118 format specifier, # construct a Numpy dtype

TypeError: Cannot change data-type for object array If I try and read in the array contained in this npy file: https://gist.github.com/astrofrog/8c2d188005f31e0bba36/raw/3065c8fa220a6eaccbff20565d0d520c07e5e7e6/test.npy then try and print out the array, so: ``` python import numpy as np array = np.load('test.npy') print(array) ``` I get: ``` Traceback (most recent call last): File "test2.py", line 5, in <module> print(array) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/numeric.py", line 1767, in array_str return array2string(a, max_line_width, precision, suppress_small, ' ', "", str) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 459, in array2string separator, prefix, formatter=formatter) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 329, in _array2string _summaryEdgeItems, summary_insert)[:-1] File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 526, in _formatArray s += _formatArray(a[-i], format_function, rank-1, max_line_len, File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 481, in __getitem__ return obj.view(dtype=(self.dtype.type, obj.dtype.descr)) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 540, in view return ndarray.view(self, dtype) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 457, in __setattr__ raise exctype(value) TypeError: Cannot change data-type for object array. ``` This is with the latest developer version of Numpy (3c5409e4e38e6034d69d0042bf2a3bc854ef2e53) and Python 3.4 on MacOS X. The dtype can be printed, as can individual columns, but the array as a whole can't. This doesn't occur in the latest stable release so may be a regression?

2015-02-09T05:49:41Z

[]

Traceback (most recent call last): File "test2.py", line 5, in <module> print(array) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/numeric.py", line 1767, in array_str return array2string(a, max_line_width, precision, suppress_small, ' ', "", str) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 459, in array2string separator, prefix, formatter=formatter) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 329, in _array2string _summaryEdgeItems, summary_insert)[:-1] File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/arrayprint.py", line 526, in _formatArray s += _formatArray(a[-i], format_function, rank-1, max_line_len, File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 481, in __getitem__ return obj.view(dtype=(self.dtype.type, obj.dtype.descr)) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 540, in view return ndarray.view(self, dtype) File "/Volumes/Raptor/miniconda3/envs/dev/lib/python3.4/site-packages/numpy/core/records.py", line 457, in __setattr__ raise exctype(value) TypeError: Cannot change data-type for object array.

10,357

numpy/numpy

numpy__numpy-5584

2e016ac65aceab4e08217794d6be7b365793976a

diff --git a/numpy/core/fromnumeric.py b/numpy/core/fromnumeric.py --- a/numpy/core/fromnumeric.py +++ b/numpy/core/fromnumeric.py @@ -691,8 +691,16 @@ def argpartition(a, kth, axis=-1, kind='introselect', order=None): >>> x[np.argpartition(x, (1, 3))] array([1, 2, 3, 4]) + >>> x = [3, 4, 2, 1] + >>> np.array(x)[np.argpartition(x, 3)] + array([2, 1, 3, 4]) + """ - return a.argpartition(kth, axis, kind=kind, order=order) + try: + argpartition = a.argpartition + except AttributeError: + return _wrapit(a, 'argpartition',kth, axis, kind, order) + return argpartition(kth, axis, kind=kind, order=order) def sort(a, axis=-1, kind='quicksort', order=None):

argpartition fails on non-ndarray array-likes While `partition` works, as advertised, array-likeinputs like, e.g. lists: ``` >>> np.partition([5, 4, 3, 2, 1], 2) array([1, 2, 3, 4, 5]) ``` `argpartition` raises an error if its first argument is not an `ndarray`: ``` >>> np.argpartition([5, 4, 3, 2, 1], 2) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition' ```

If any array-like inputs are to be accepted, then as in `argsort`, this can be implemented: ``` try: argpartition = a.argpartition except AttributeError: return _wrapit(a, 'argpartition', kth, axis, kind, order) return argpartition(kth, axis, kind=kind, order=order) ``` else it can be like `sort` and `partition`: ``` if axis is None: a = asanyarray(a).flatten() axis = 0 else: a = asanyarray(a).copy(order="K") return a.argpartition(kth, axis, kind=kind, order=order) ``` Just asking out of curiosity, the following example is shown in the documentation : ``` >>> x = np.array([3, 4, 2, 1]) >>> x[np.argpartition(x, 3)] array([2, 1, 3, 4]) ``` Now, if any array-like inputs are allowed, then the following behavior is needed. ``` >>> x = [3, 4, 2, 1] >>>np.array(x)[np.argpartition(x, 3)] array([2, 1, 3, 4]) >>>x[np.argpartition(x, 3)] Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: only integer arrays with one element can be converted to an index ``` If this is the procedure to be followed, as in case of `argsort`, I think it should be documented. Please correct me if no such use is needed in general scenarios. it probably should be like argsort, Don't remember why its not, maybe just a mistake, I can't remember there being a reason. compared to sort, partition can't be overridden from the C side but that shouldn't matter here. I think the `partition` and `sort` C API functions are defined in `item_selection.c`. Can you please clarify on what `compared to sort, partition can't be overridden from the C side` means ? Could I have a go at fixing this ? no, that is kind of intentional the C sort functions are part of a public struct, adding partition would break the ABI and just add more stuff that should not be in that place. So I went with a internal lookupt able that could get hooks for user functions. But I don't think its really worthwhile, I don't think anybody would use it. @juliantaylor Thanks for the clarification. I am new to working with these kind of projects and wasn't aware of the nature of public struct functions and the way in which they are maintained.

2015-02-19T13:56:08Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition'

10,359

numpy/numpy

numpy__numpy-5616

bf6f80d994a154f25eb5e8beca5babdf31b13eca

diff --git a/numpy/core/fromnumeric.py b/numpy/core/fromnumeric.py --- a/numpy/core/fromnumeric.py +++ b/numpy/core/fromnumeric.py @@ -679,8 +679,16 @@ def argpartition(a, kth, axis=-1, kind='introselect', order=None): >>> x[np.argpartition(x, (1, 3))] array([1, 2, 3, 4]) + >>> x = [3, 4, 2, 1] + >>> np.array(x)[np.argpartition(x, 3)] + array([2, 1, 3, 4]) + """ - return a.argpartition(kth, axis, kind=kind, order=order) + try: + argpartition = a.argpartition + except AttributeError: + return _wrapit(a, 'argpartition',kth, axis, kind, order) + return argpartition(kth, axis, kind=kind, order=order) def sort(a, axis=-1, kind='quicksort', order=None):

argpartition fails on non-ndarray array-likes While `partition` works, as advertised, array-likeinputs like, e.g. lists: ``` >>> np.partition([5, 4, 3, 2, 1], 2) array([1, 2, 3, 4, 5]) ``` `argpartition` raises an error if its first argument is not an `ndarray`: ``` >>> np.argpartition([5, 4, 3, 2, 1], 2) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition' ```

If any array-like inputs are to be accepted, then as in `argsort`, this can be implemented: ``` try: argpartition = a.argpartition except AttributeError: return _wrapit(a, 'argpartition', kth, axis, kind, order) return argpartition(kth, axis, kind=kind, order=order) ``` else it can be like `sort` and `partition`: ``` if axis is None: a = asanyarray(a).flatten() axis = 0 else: a = asanyarray(a).copy(order="K") return a.argpartition(kth, axis, kind=kind, order=order) ``` Just asking out of curiosity, the following example is shown in the documentation : ``` >>> x = np.array([3, 4, 2, 1]) >>> x[np.argpartition(x, 3)] array([2, 1, 3, 4]) ``` Now, if any array-like inputs are allowed, then the following behavior is needed. ``` >>> x = [3, 4, 2, 1] >>>np.array(x)[np.argpartition(x, 3)] array([2, 1, 3, 4]) >>>x[np.argpartition(x, 3)] Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: only integer arrays with one element can be converted to an index ``` If this is the procedure to be followed, as in case of `argsort`, I think it should be documented. Please correct me if no such use is needed in general scenarios. it probably should be like argsort, Don't remember why its not, maybe just a mistake, I can't remember there being a reason. compared to sort, partition can't be overridden from the C side but that shouldn't matter here. I think the `partition` and `sort` C API functions are defined in `item_selection.c`. Can you please clarify on what `compared to sort, partition can't be overridden from the C side` means ? Could I have a go at fixing this ? no, that is kind of intentional the C sort functions are part of a public struct, adding partition would break the ABI and just add more stuff that should not be in that place. So I went with a internal lookupt able that could get hooks for user functions. But I don't think its really worthwhile, I don't think anybody would use it. @juliantaylor Thanks for the clarification. I am new to working with these kind of projects and wasn't aware of the nature of public struct functions and the way in which they are maintained.

2015-02-28T13:04:47Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\WinPython-32bit-sp-dev\python-2.7.5\lib\site-packages\numpy\core\fromnumeric.py", line 680, in argpartition return a.argpartition(kth, axis, kind=kind, order=order) AttributeError: 'list' object has no attribute 'argpartition'

10,361

numpy/numpy

numpy__numpy-5742

147c60f83f401037ff29593826d2c5729a73c2c5

diff --git a/numpy/lib/type_check.py b/numpy/lib/type_check.py --- a/numpy/lib/type_check.py +++ b/numpy/lib/type_check.py @@ -324,12 +324,13 @@ def nan_to_num(x): Returns ------- - out : ndarray, float - Array with the same shape as `x` and dtype of the element in `x` with - the greatest precision. NaN is replaced by zero, and infinity - (-infinity) is replaced by the largest (smallest or most negative) - floating point value that fits in the output dtype. All finite numbers - are upcast to the output dtype (default float64). + out : ndarray + New Array with the same shape as `x` and dtype of the element in + `x` with the greatest precision. If `x` is inexact, then NaN is + replaced by zero, and infinity (-infinity) is replaced by the + largest (smallest or most negative) floating point value that fits + in the output dtype. If `x` is not inexact, then a copy of `x` is + returned. See Also -------- @@ -354,33 +355,22 @@ def nan_to_num(x): -1.28000000e+002, 1.28000000e+002]) """ - try: - t = x.dtype.type - except AttributeError: - t = obj2sctype(type(x)) - if issubclass(t, _nx.complexfloating): - return nan_to_num(x.real) + 1j * nan_to_num(x.imag) - else: - try: - y = x.copy() - except AttributeError: - y = array(x) - if not issubclass(t, _nx.integer): - if not y.shape: - y = array([x]) - scalar = True - else: - scalar = False - are_inf = isposinf(y) - are_neg_inf = isneginf(y) - are_nan = isnan(y) - maxf, minf = _getmaxmin(y.dtype.type) - y[are_nan] = 0 - y[are_inf] = maxf - y[are_neg_inf] = minf - if scalar: - y = y[0] - return y + x = _nx.array(x, subok=True) + xtype = x.dtype.type + if not issubclass(xtype, _nx.inexact): + return x + + iscomplex = issubclass(xtype, _nx.complexfloating) + isscalar = (x.ndim == 0) + + x = x[None] if isscalar else x + dest = (x.real, x.imag) if iscomplex else (x,) + maxf, minf = _getmaxmin(x.real.dtype) + for d in dest: + _nx.copyto(d, 0.0, where=isnan(d)) + _nx.copyto(d, maxf, where=isposinf(d)) + _nx.copyto(d, minf, where=isneginf(d)) + return x[0] if isscalar else x #-----------------------------------------------------------------------------

An error is given if a integer list is passed to nan_to_num (Trac #880) _Original ticket http://projects.scipy.org/numpy/ticket/880 on 2008-08-06 by @bsouthey, assigned to unknown._ When an integer list is passed to nan_to_num then an error occurs. This does not occur if the list contains floats or a numpy integer array is used. ``` np.nan_to_num([1.0,3]) # returns: array([ 1., 3.]) n=np.array([1,3]) np.nan_to_num(n) # returns: array([1, 3]) np.nan_to_num([1,3]) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 135, in nan_to_num maxf, minf = _getmaxmin(y.dtype.type) File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 103, in _getmaxmin f = getlimits.finfo(t) File "/usr/lib64/python2.5/site-packages/numpy/lib/getlimits.py", line 46, in __new__ raise ValueError, "data type %r not inexact" % (dtype) ValueError: data type <type 'numpy.int64'> not inexact ```

Milestone changed to `1.4.0` by @cournape on 2009-03-09 Milestone changed to `Unscheduled` by @mwiebe on 2011-03-23 Still present in 1.9-devel.

2015-04-03T17:34:50Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 135, in nan_to_num maxf, minf = _getmaxmin(y.dtype.type) File "/usr/lib64/python2.5/site-packages/numpy/lib/type_check.py", line 103, in _getmaxmin f = getlimits.finfo(t) File "/usr/lib64/python2.5/site-packages/numpy/lib/getlimits.py", line 46, in __new__ raise ValueError, "data type %r not inexact" % (dtype) ValueError: data type <type 'numpy.int64'> not inexact

10,367

numpy/numpy

numpy__numpy-6556

2d899ea2301e155b06acf585020866ca1953bce5

diff --git a/numpy/lib/stride_tricks.py b/numpy/lib/stride_tricks.py --- a/numpy/lib/stride_tricks.py +++ b/numpy/lib/stride_tricks.py @@ -62,11 +62,14 @@ def _broadcast_to(array, shape, subok, readonly): if any(size < 0 for size in shape): raise ValueError('all elements of broadcast shape must be non-' 'negative') + needs_writeable = not readonly and array.flags.writeable + extras = ['reduce_ok'] if needs_writeable else [] + op_flag = 'readwrite' if needs_writeable else 'readonly' broadcast = np.nditer( - (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'], - op_flags=['readonly'], itershape=shape, order='C').itviews[0] + (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'] + extras, + op_flags=[op_flag], itershape=shape, order='C').itviews[0] result = _maybe_view_as_subclass(array, broadcast) - if not readonly and array.flags.writeable: + if needs_writeable and not result.flags.writeable: result.flags.writeable = True return result

Error in broadcasting stride_tricks array We hit a new test failure when testing against numpy 1.10.1 : https://github.com/nipy/nibabel/pull/358 The test failure boils down to this: ``` import numpy as np shape = (2,) strides = [0] tricky_scalar = np.lib.stride_tricks.as_strided(np.array(0), shape, strides) other = np.zeros((1,)) first, second = np.broadcast_arrays(tricky_scalar, other) ``` On Linux (not OSX) I get the following error, for numpy 1.10.1: ``` Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array ``` There is no error for the same code on numpy 1.9.3. I think this is the same issue, arising in scipy: https://github.com/scipy/scipy/pull/5374#issuecomment-148775998

Hmm. I wrote the offending line of code, so I can take a look at this. @shoyer Any progress?

2015-10-24T19:01:58Z

[]

Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array

10,403

numpy/numpy

numpy__numpy-6557

91093ff8c1ad9ccf8096fadc2e695d4039e529fc

diff --git a/numpy/lib/stride_tricks.py b/numpy/lib/stride_tricks.py --- a/numpy/lib/stride_tricks.py +++ b/numpy/lib/stride_tricks.py @@ -62,11 +62,14 @@ def _broadcast_to(array, shape, subok, readonly): if any(size < 0 for size in shape): raise ValueError('all elements of broadcast shape must be non-' 'negative') + needs_writeable = not readonly and array.flags.writeable + extras = ['reduce_ok'] if needs_writeable else [] + op_flag = 'readwrite' if needs_writeable else 'readonly' broadcast = np.nditer( - (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'], - op_flags=['readonly'], itershape=shape, order='C').itviews[0] + (array,), flags=['multi_index', 'refs_ok', 'zerosize_ok'] + extras, + op_flags=[op_flag], itershape=shape, order='C').itviews[0] result = _maybe_view_as_subclass(array, broadcast) - if not readonly and array.flags.writeable: + if needs_writeable and not result.flags.writeable: result.flags.writeable = True return result

Error in broadcasting stride_tricks array We hit a new test failure when testing against numpy 1.10.1 : https://github.com/nipy/nibabel/pull/358 The test failure boils down to this: ``` import numpy as np shape = (2,) strides = [0] tricky_scalar = np.lib.stride_tricks.as_strided(np.array(0), shape, strides) other = np.zeros((1,)) first, second = np.broadcast_arrays(tricky_scalar, other) ``` On Linux (not OSX) I get the following error, for numpy 1.10.1: ``` Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array ``` There is no error for the same code on numpy 1.9.3. I think this is the same issue, arising in scipy: https://github.com/scipy/scipy/pull/5374#issuecomment-148775998

Hmm. I wrote the offending line of code, so I can take a look at this. @shoyer Any progress?

2015-10-24T20:53:18Z

[]

Traceback (most recent call last): File "broadcast_bug.py", line 7, in <module> first, second = np.broadcast_arrays(tricky_scalar, other) File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 200, in broadcast_arrays for array in args] File "/home/mb312/.virtualenvs/test/local/lib/python2.7/site-packages/numpy/lib/stride_tricks.py", line 70, in _broadcast_to result.flags.writeable = True ValueError: cannot set WRITEABLE flag to True of this array

10,404

numpy/numpy

numpy__numpy-6905

e072d79f03610c33e336a9b700882d8905f9c958

diff --git a/numpy/lib/stride_tricks.py b/numpy/lib/stride_tricks.py --- a/numpy/lib/stride_tricks.py +++ b/numpy/lib/stride_tricks.py @@ -121,9 +121,6 @@ def _broadcast_shape(*args): """ if not args: raise ValueError('must provide at least one argument') - if len(args) == 1: - # a single argument does not work with np.broadcast - return np.asarray(args[0]).shape # use the old-iterator because np.nditer does not handle size 0 arrays # consistently b = np.broadcast(*args[:32])

Why does `numpy.broadcast` not accept a single input array? I'm wondering what's the reasoning behind forbidding a single array as input to `numpy.broadcast`. When writing code which is supposed to work in N dimensions with N >= 1, it would be nice to be able to use `broadcast` as-is. A standard use case is the calculation of the shape of an output array when using meshgrids as input for function evaluation: ``` python >>> coord_vecs = [[1, 2], [3, 4, 5]] >>> mesh = np.meshgrid(*coord_vecs, indexing='ij', sparse=True) # for large stuff >>> out_shape = np.broadcast(*mesh).shape >>> out_shape (2, 3) >>> coord_vecs = [[1, 2]] >>> mesh = np.meshgrid(*coord_vecs, indexing='ij', sparse=True) # still works >>> out_shape = np.broadcast(*mesh).shape Traceback (most recent call last): ... ValueError: Need at least two and fewer than (32) array objects. ``` As noted above, `meshgrid` works and returns a list with one element, as expected. Why does `broadcast` not return a broadcast object with shape `(2,)` in the example above? The other public methods of such an object would probably also make sense with a single input array. EDIT: Last sentence was wrong in the first version.

Apparently, there is [one single place](https://github.com/numpy/numpy/blob/004639d07fd161d1394f5dda1b6ed42c777f3c80/numpy/core/src/multiarray/iterators.c#L1606-L1614) where this situation is caught. Without knowing much about internals, the function around that check seems to be completely generic, and simply replacing 2 with 1 could be the only change necessary to allow single-array(-like) input. No idea about side effects of such a change, though. Well, maybe worth some playing around, then? I also find this surprising. Seems to work right away with the proposed (trivial) change: ``` python >>> import numpy as np >>> np.__path__ # My local GH copy ['/home/hkohr/Software/numpy/numpy'] >>> a = np.arange(3) >>> bc = np.broadcast(a) >>> bc.nd 1 >>> bc.numiter 1 >>> bc.shape (3,) >>> bc.size 3 ``` I get three errors when running `np.test()`, but they're probably unrelated. I just post them for the record: ``` ====================================================================== ERROR: Failure: ImportError (cannot import name ccompiler) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/loader.py", line 418, in loadTestsFromName addr.filename, addr.module) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/importer.py", line 47, in importFromPath return self.importFromDir(dir_path, fqname) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/importer.py", line 94, in importFromDir mod = load_module(part_fqname, fh, filename, desc) File "numpy/distutils/__init__.py", line 8, in <module> from . import ccompiler File "numpy/distutils/ccompiler.py", line 8, in <module> from distutils import ccompiler File "numpy/distutils/__init__.py", line 8, in <module> from . import ccompiler File "numpy/distutils/ccompiler.py", line 8, in <module> from distutils import ccompiler ImportError: cannot import name ccompiler ====================================================================== ERROR: test suite for <module 'test_array_from_pyobj' from '/home/hkohr/Software/numpy/numpy/f2py/tests/test_array_from_pyobj.py'> ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/suite.py", line 209, in run self.setUp() File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/suite.py", line 292, in setUp self.setupContext(ancestor) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/suite.py", line 315, in setupContext try_run(context, names) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/util.py", line 471, in try_run return func() File "/home/hkohr/Software/numpy/numpy/f2py/tests/test_array_from_pyobj.py", line 42, in setup 'test_array_from_pyobj_ext') File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 78, in wrapper memo[key] = func(*a, **kw) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 312, in build_module_distutils __import__(module_name) ImportError: /tmp/tmpqyEUs_/test_array_from_pyobj_ext.so: failed to map segment from shared object ====================================================================== ERROR: test_callback.TestF77Callback.test_string_callback ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/case.py", line 381, in setUp try_run(self.inst, ('setup', 'setUp')) File "/home/hkohr/.virtualenvs/numpy-py2.7/lib/python2.7/site-packages/nose/util.py", line 471, in try_run return func() File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 361, in setUp module_name=self.module_name) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 78, in wrapper memo[key] = func(*a, **kw) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 169, in build_code module_name=module_name) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 78, in wrapper memo[key] = func(*a, **kw) File "/home/hkohr/Software/numpy/numpy/f2py/tests/util.py", line 149, in build_module __import__(module_name) ImportError: /tmp/tmpqyEUs_/_test_ext_module_5403.so: failed to map segment from shared object ---------------------------------------------------------------------- ``` +1 on exploring this change: looking at the code it certainly seems like an unnecessary restriction. Your test errors do not seem related, do you get the same errors if testing master unchanged? Would be good to put a PR together and see what Travis thinks.

2015-12-30T12:09:25Z

[]

Traceback (most recent call last): ... ValueError: Need at least two and fewer than (32) array objects.

10,428

numpy/numpy

numpy__numpy-7133

e2805398f9a63b825f4a2aab22e9f169ff65aae9

diff --git a/numpy/lib/npyio.py b/numpy/lib/npyio.py --- a/numpy/lib/npyio.py +++ b/numpy/lib/npyio.py @@ -627,7 +627,11 @@ def _savez(file, args, kwds, compress, allow_pickle=True, pickle_kwargs=None): zipf = zipfile_factory(file, mode="w", compression=compression) # Stage arrays in a temporary file on disk, before writing to zip. - fd, tmpfile = tempfile.mkstemp(suffix='-numpy.npy') + + # Since target file might be big enough to exceed capacity of a global + # temporary directory, create temp file side-by-side with the target file. + file_dir, file_prefix = os.path.split(file) if _is_string_like(file) else (None, 'tmp') + fd, tmpfile = tempfile.mkstemp(prefix=file_prefix, dir=file_dir, suffix='-numpy.npy') os.close(fd) try: for key, val in namedict.items(): @@ -640,6 +644,8 @@ def _savez(file, args, kwds, compress, allow_pickle=True, pickle_kwargs=None): fid.close() fid = None zipf.write(tmpfile, arcname=fname) + except IOError as exc: + raise IOError("Failed to write to %s: %s" % (tmpfile, exc)) finally: if fid: fid.close()

Saving large array fails with savez_compressed() but works with save() I got this strange problem on a CentOS box, Python 2.6.6, Numpy 1.9.1: ``` [minhle@node069 ~]$ python Python 2.6.6 (r266:84292, Jan 22 2014, 09:42:36) [GCC 4.4.7 20120313 (Red Hat 4.4.7-4)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import numpy as np >>> np.version.version '1.9.1' >>> a = np.ones((775890380,)) >>> np.savez_compressed('/home/minhle/scratch/test.npz', a) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 560, in savez_compressed _savez(file, args, kwds, True) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 597, in _savez format.write_array(fid, np.asanyarray(val)) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/format.py", line 562, in write_array array.tofile(fp) IOError: 775890380 requested and 233691638 written >>> np.save('/home/minhle/scratch/test.npy', a) >>> b = np.load('/home/minhle/scratch/test.npy') >>> b[:10] array([ 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.]) >>> quit() [minhle@node069 ~]$ cat /etc/*-release CentOS release 6.5 (Final) Cluster Manager v5.2 slave LSB_VERSION=base-4.0-amd64:base-4.0-noarch:core-4.0-amd64:core-4.0-noarch:graphics-4.0-amd64:graphics-4.0-noarch:printing-4.0-amd64:printing-4.0-noarch CentOS release 6.5 (Final) CentOS release 6.5 (Final) ```

Got the same error... any solution in sight? I just got this error too but with `np.savez` as opposed to `np.savez_compressed`. For future reference, using `np.save` instead of `np.savez[_compressed]` seems to work as @omerlevy's example code indicates. So, is this an issue with `savez`? Anyone who knows numpy well have any ideas what's happening here? `savez` needs to store files temporarily on disk, and perhaps you run out of space on `/tmp`. Try setting `TMPDIR=$HOME/tmp` or so. @pv Yeah, changing `TMPDIR` did the trick for me. It may also be worth noting that I'm on an NFS filesystem. I'm not sure what sort of effects, if any, that might have. I'm guessing that I'm into subtleties in the way that /tmp is configured. However, I think numpy should make this issue explicit in the documentation and provide some semblance of an intelligent, approachable error message. Agreed. I ran onto this issue on HPC resources too and I agree that it would be good if the error / warning message recommended checking to make sure `TMPDIR` was set properly with adequate space. Ran into the same problem (kudos to ngoldbaum@IRC for pointing me here). I understand that for temp storage TEMPDIR is usually the logical location. BUT when dealing with large datasets, there is a reason why we might want to savez to a different location and /tmp would never be big enough. So why not to follow any other downloader's behavior and use targetfilename + '_tempSMTHRANDOM' suffix instead. This has higher possibility to succeed since would use the target location/partition which should have adequate amount of storage. Also error message should include location of that temp file to give users sensible feedback

2016-01-28T03:16:40Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 560, in savez_compressed _savez(file, args, kwds, True) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/npyio.py", line 597, in _savez format.write_array(fid, np.asanyarray(val)) File "/home/minhle/.local/lib/python2.6/site-packages/numpy/lib/format.py", line 562, in write_array array.tofile(fp) IOError: 775890380 requested and 233691638 written

10,431

numpy/numpy

numpy__numpy-7152

9cfdb218b16cba97800fc2ad0f37e1af780ab664

diff --git a/numpy/lib/arraypad.py b/numpy/lib/arraypad.py --- a/numpy/lib/arraypad.py +++ b/numpy/lib/arraypad.py @@ -1337,7 +1337,7 @@ def pad(array, pad_width, mode, **kwargs): 'reflect_type': 'even', } - if isinstance(mode, str): + if isinstance(mode, np.compat.basestring): # Make sure have allowed kwargs appropriate for mode for key in kwargs: if key not in allowedkwargs[mode]:

from __future__ import unicode_literals breaks numpy.pad The `pad` function accepts a `mode` parameter which can be a callable or str. [It uses the test `isinstance(mode, str)` to determine which is the case](https://github.com/numpy/numpy/blob/master/numpy/lib/arraypad.py#L1340). Using `from __future__ import unicode_literals` makes this fail, as the `isinstance` call returns `False`, eventually producing a `TypeError: 'unicode' object is not callable` when the `mode` argument is instead called as a function. For example, the script: ``` from __future__ import unicode_literals import numpy as np np.pad([10], 2, mode='constant') ``` Fails using `python2.7` with the error: ``` Traceback (most recent call last): File "test.py", line 4, in <module> np.pad([10], 2, mode='constant') File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/arraypad.py", line 1348, in pad kwargs) File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/shape_base.py", line 79, in apply_along_axis res = func1d(arr[tuple(i.tolist())],*args) TypeError: 'unicode' object is not callable ``` --- Would using the [`six.string_types`](https://pythonhosted.org/six/#six.string_types) comparison [_a la_ this stackexchange post](http://stackoverflow.com/a/11301392/230468) be a suitable fix?

Yeah, `np.pad` (and in general anything that takes a string `mode`-style argument) ought to accept unicode strings on py2. I don't know what the very simplest cleanest way to write this check would be, but we'd certainly accept a pull request fixing it. (Maybe a helper function like ``` def _py2_fixup_unicode(s): if this_is_py2 and isinstance(s, unicode): try: return s.encode("ascii") except UnicodeEncodeError: pass return s ``` and then functions with `mode`-style arguments do `mode = _fixup_py2_unicode(mode)` at the top?) It sounds like that might be the best, most-general solution. In which case, is `sys.version_info` a reliable way to check the python version? For `numpy.pad` in particular, since the only options are string and callable, what about using ``` if callable(mode): # treat as function.... else: # treat as str ``` That could also work, though since the specific thing we're worrying about is unicode-on-py2 I kinda prefer checking for unicode -- plus in your way, the code below has to be prepared to handle both unicode and str, while in mine it only has to work for str. But really there are lots of ways that would work -- someone just needs to write the patch :-) On Sun, Jan 24, 2016 at 2:50 PM, Luke notifications@github.com wrote: > It sounds like that might be the best, most-general solution. In which > case, is sys.version_info a reliable way to check the python version? > > For numpy.pad in particular, since the only options are string and > callable, what about using > > if callable(mode): > # treat as function.... > else: > # treat as str > > — > Reply to this email directly or view it on GitHub > https://github.com/numpy/numpy/issues/7112#issuecomment-174349793. ## Nathaniel J. Smith -- https://vorpus.org http://vorpus.org `np.compat.basestring` and/or `np.compat.sixu` might help? Anyway, I am not sure whether `np.pad` may be the least of our problems, do array fields and other C-side string arguments work with it? I don't remember fixing it, and I seem to remember it was a problem. Now that we have dropped Python 3.2 and 3.3, sixu can be replaced by `u"..."`. Unicode literals broke a lot of code, which is why they weren't part of the Python2, Python3 code unification, but it may be that we can do something about that now. I don't recall the exact problems that came up. The issue here isn't that we need Unicode literals but that np.pad should respond gracefully if someone passes in a Unicode literal. I agree that there are probably other places in the code that don't handle this well either (I think that's @seberg's point?), but I guess we should wait for a bug to be reported and then fix those too? On Jan 25, 2016 7:09 AM, "Charles Harris" notifications@github.com wrote: > Now that we have dropped Python 3.2 and 3.3, sixu can be replaced by > u"...". Unicode literals broke a lot of code, which is why they weren't > part of the Python2, Python3 code unification, but it may be that we can do > something about that now. I don't recall the exact problems that came up. > > — > Reply to this email directly or view it on GitHub > https://github.com/numpy/numpy/issues/7112#issuecomment-174537453. Yeah, as far as I understand Chuck we might now have a better chance to fix things then before. Array fields in dtypes definitely don't work in all regards at least. But that does not mean we should not just fix this. If we ever find something we cannot fix then so be it, but this should not be hard to fix.

2016-01-31T02:13:23Z

[]

Traceback (most recent call last): File "test.py", line 4, in <module> np.pad([10], 2, mode='constant') File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/arraypad.py", line 1348, in pad kwargs) File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/shape_base.py", line 79, in apply_along_axis res = func1d(arr[tuple(i.tolist())],*args) TypeError: 'unicode' object is not callable

10,432

numpy/numpy

numpy__numpy-7587

1fc180b4c683e79649e5699303722995ca3e8ef9

diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -23,7 +23,7 @@ csingle, cdouble, inexact, complexfloating, newaxis, ravel, all, Inf, dot, add, multiply, sqrt, maximum, fastCopyAndTranspose, sum, isfinite, size, finfo, errstate, geterrobj, longdouble, rollaxis, amin, amax, product, abs, - broadcast, atleast_2d, intp, asanyarray, isscalar + broadcast, atleast_2d, intp, asanyarray, isscalar, object_ ) from numpy.lib import triu, asfarray from numpy.linalg import lapack_lite, _umath_linalg @@ -2112,7 +2112,7 @@ def norm(x, ord=None, axis=None, keepdims=False): """ x = asarray(x) - if not issubclass(x.dtype.type, inexact): + if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) # Immediately handle some default, simple, fast, and common cases.

Regression in linalg.norm() using dtype=object In NumPy 1.10.1, this works: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) array([ 0., 1.]) ``` In NumPy 1.11.0, however, it raises an exception: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence. ``` Probably this ``` python if not issubclass(x.dtype.type, inexact): x = x.astype(float) ``` should be changed to ``` python if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) ``` ? Does that make sense? If yes, I can make a PR.

2016-04-29T08:51:38Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence.

10,456

numpy/numpy

numpy__numpy-7608

730e2881219f4af79ab0ad5930a3bea5ab60b098

diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -23,7 +23,7 @@ csingle, cdouble, inexact, complexfloating, newaxis, ravel, all, Inf, dot, add, multiply, sqrt, maximum, fastCopyAndTranspose, sum, isfinite, size, finfo, errstate, geterrobj, longdouble, rollaxis, amin, amax, product, abs, - broadcast, atleast_2d, intp, asanyarray, isscalar + broadcast, atleast_2d, intp, asanyarray, isscalar, object_ ) from numpy.lib import triu, asfarray from numpy.linalg import lapack_lite, _umath_linalg @@ -2112,7 +2112,7 @@ def norm(x, ord=None, axis=None, keepdims=False): """ x = asarray(x) - if not issubclass(x.dtype.type, inexact): + if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) # Immediately handle some default, simple, fast, and common cases.

Regression in linalg.norm() using dtype=object In NumPy 1.10.1, this works: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) array([ 0., 1.]) ``` In NumPy 1.11.0, however, it raises an exception: ``` python >>> import numpy as np >>> np.linalg.norm(np.array([np.array([0, 1]), 0, 0], dtype=object)) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence. ``` Probably this ``` python if not issubclass(x.dtype.type, inexact): x = x.astype(float) ``` should be changed to ``` python if not issubclass(x.dtype.type, (inexact, object_)): x = x.astype(float) ``` ? Does that make sense? If yes, I can make a PR.

2016-05-07T01:56:25Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3/dist-packages/numpy/linalg/linalg.py", line 2116, in norm x = x.astype(float) ValueError: setting an array element with a sequence.

10,458

numpy/numpy

numpy__numpy-8148

b0a6e5583f2952a040c45d0d50f5e93372dc175b

diff --git a/numpy/compat/py3k.py b/numpy/compat/py3k.py --- a/numpy/compat/py3k.py +++ b/numpy/compat/py3k.py @@ -118,7 +118,7 @@ def npy_load_module(name, fn, info=None): mod : module """ - import importlib + import importlib.machinery return importlib.machinery.SourceFileLoader(name, fn).load_module() else: def npy_load_module(name, fn, info=None):

Configuration.add_subpackage fails on Python 3.4 Using numpy 1.11.2 and Python 3.4, trying to run a `setupy.py` script which uses subpackage using numpy.distutil fails when `importlib.machinery` cannot be found in the `npy_load_module` function. For example using the following `setup.py` and `foo/setup.py` scripts: ``` Python # setup.py from numpy.distutils.misc_util import Configuration from numpy.distutils.core import setup def configuration(): config = Configuration(None, '', None) config.add_subpackage('foo') return config setup(configuration=configuration) ``` ``` Python # foo/setup.py def configuration(parent_package='', top_path=None): from numpy.distutils.misc_util import Configuration config = Configuration('foo', parent_package, top_path) return config ``` Attempting to build the package: ``` $ python setup.py build_ext -i Traceback (most recent call last): File "setup.py", line 10, in <module> setup(configuration=configuration) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/core.py", line 135, in setup config = configuration() File "setup.py", line 7, in configuration config.add_subpackage('foo') File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 1000, in add_subpackage caller_level = 2) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 969, in get_subpackage caller_level = caller_level + 1) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 882, in _get_configuration_from_setup_py ('.py', 'U', 1)) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/compat/py3k.py", line 112, in npy_load_module return importlib.machinery.SourceFileLoader(name, fn).load_module() AttributeError: 'module' object has no attribute 'machinery' ``` This works in Python 3.5 and 2.7 and with earlier versions of NumPy prior to the addition of `npy_load_module`. This bug can likely be fixed by adding `import importlib` to the `npy_load_module`. I will submit a PR shortly. A workaround is to `import importlib.machinery` somewhere in the root `setup.py` file.

2016-10-12T18:13:58Z

[]

Traceback (most recent call last): File "setup.py", line 10, in <module> setup(configuration=configuration) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/core.py", line 135, in setup config = configuration() File "setup.py", line 7, in configuration config.add_subpackage('foo') File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 1000, in add_subpackage caller_level = 2) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 969, in get_subpackage caller_level = caller_level + 1) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/distutils/misc_util.py", line 882, in _get_configuration_from_setup_py ('.py', 'U', 1)) File "/Users/jhelmus/anaconda/envs/py34/lib/python3.4/site-packages/numpy/compat/py3k.py", line 112, in npy_load_module return importlib.machinery.SourceFileLoader(name, fn).load_module() AttributeError: 'module' object has no attribute 'machinery'

10,492

numpy/numpy

numpy__numpy-8384

5f5ccecbfc116284ed8c8d53cd8b203ceef5f7c7

diff --git a/numpy/core/code_generators/genapi.py b/numpy/core/code_generators/genapi.py --- a/numpy/core/code_generators/genapi.py +++ b/numpy/core/code_generators/genapi.py @@ -469,7 +469,7 @@ def fullapi_hash(api_dicts): # To parse strings like 'hex = checksum' where hex is e.g. 0x1234567F and # checksum a 128 bits md5 checksum (hex format as well) -VERRE = re.compile('(^0x[\da-f]{8})\s*=\s*([\da-f]{32})') +VERRE = re.compile(r'(^0x[\da-f]{8})\s*=\s*([\da-f]{32})') def get_versions_hash(): d = [] diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -1644,7 +1644,7 @@ def gradient(f, *varargs, **kwargs): + \\left(h_{d}^{2} - h_{s}^{2}\\right)f\\left(x_{i}\\right) - h_{d}^{2}f\\left(x_{i}-h_{s}\\right)} { h_{s}h_{d}\\left(h_{d} + h_{s}\\right)} - + \mathcal{O}\\left(\\frac{h_{d}h_{s}^{2} + + \\mathcal{O}\\left(\\frac{h_{d}h_{s}^{2} + h_{s}h_{d}^{2}}{h_{d} + h_{s}}\\right) @@ -1656,7 +1656,7 @@ def gradient(f, *varargs, **kwargs): \\hat f_{i}^{(1)}= \\frac{f\\left(x_{i+1}\\right) - f\\left(x_{i-1}\\right)}{2h} - + \mathcal{O}\\left(h^{2}\\right) + + \\mathcal{O}\\left(h^{2}\\right) With a similar procedure the forward/backward approximations used for boundaries can be derived. diff --git a/tools/find_deprecated_escaped_characters.py b/tools/find_deprecated_escaped_characters.py new file mode 100644 --- /dev/null +++ b/tools/find_deprecated_escaped_characters.py @@ -0,0 +1,69 @@ +#! /usr/bin/env python +""" +Look for escape sequences deprecated in Python 3.6. + +Python 3.6 deprecates a number of non-escape sequences starting with `\` that +were accepted before. For instance, '\(' was previously accepted but must now +be written as '\\(' or r'\('. + +""" +from __future__ import division, absolute_import, print_function + +import sys + +def main(root): + """Find deprecated escape sequences. + + Checks for deprecated escape sequences in ``*.py files``. If `root` is a + file, that file is checked, if `root` is a directory all ``*.py`` files + found in a recursive descent are checked. + + If a deprecated escape sequence is found, the file and line where found is + printed. Note that for multiline strings the line where the string ends is + printed and the error(s) are somewhere in the body of the string. + + Parameters + ---------- + root : str + File or directory to check. + Returns + ------- + None + + """ + count = 0 + + if sys.version_info[:2] >= (3, 6): + import ast + import tokenize + import warnings + from pathlib import Path + + base = Path(root) + paths = base.rglob("*.py") if base.is_dir() else [base] + for path in paths: + # use tokenize to auto-detect encoding on systems where no + # default encoding is defined (e.g. LANG='C') + with tokenize.open(str(path)) as f: + with warnings.catch_warnings(record=True) as w: + warnings.simplefilter('always') + tree = ast.parse(f.read()) + if w: + print("file: ", str(path)) + for e in w: + print('line: ', e.lineno, ': ', e.message) + print() + count += len(w) + else: + raise RuntimeError("Python version must be >= 3.6") + + print("Errors Found", count) + + +if __name__ == "__main__": + from argparse import ArgumentParser + + parser = ArgumentParser(description="Find deprecated escaped characters") + parser.add_argument('root', help='directory or file to be checked') + args = parser.parse_args() + main(args.root)

`test_warning_calls` error on Python 3.6 As of today, the daily numpy wheel builds show the following error on master for all Python 3.6 builds: ``` ====================================================================== ERROR: test_warnings.test_warning_calls ---------------------------------------------------------------------- Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 81, in test_warning_calls tree = ast.parse(file.read()) File "/opt/cp36m/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 1675 SyntaxError: invalid escape sequence \m ``` https://travis-ci.org/MacPython/numpy-wheels/builds/204888912 This is testing commit 5f5ccecbf . I see that the usual travis-ci builds tested the same commit without error : https://travis-ci.org/numpy/numpy/builds/204805650 . Any thoughts?

2016-12-14T20:14:05Z

[]

Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 81, in test_warning_calls tree = ast.parse(file.read()) File "/opt/cp36m/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 1675 SyntaxError: invalid escape sequence \m

10,504

numpy/numpy

numpy__numpy-8497

ee3ab365cb55cce6d0b9b6ed5cfbd8e3ede8cc66

diff --git a/numpy/matrixlib/defmatrix.py b/numpy/matrixlib/defmatrix.py --- a/numpy/matrixlib/defmatrix.py +++ b/numpy/matrixlib/defmatrix.py @@ -3,49 +3,15 @@ __all__ = ['matrix', 'bmat', 'mat', 'asmatrix'] import sys +import ast import numpy.core.numeric as N from numpy.core.numeric import concatenate, isscalar, binary_repr, identity, asanyarray from numpy.core.numerictypes import issubdtype -# make translation table -_numchars = '0123456789.-+jeEL' - -if sys.version_info[0] >= 3: - class _NumCharTable: - def __getitem__(self, i): - if chr(i) in _numchars: - return chr(i) - else: - return None - _table = _NumCharTable() - def _eval(astr): - str_ = astr.translate(_table) - if not str_: - raise TypeError("Invalid data string supplied: " + astr) - else: - return eval(str_) - -else: - _table = [None]*256 - for k in range(256): - _table[k] = chr(k) - _table = ''.join(_table) - - _todelete = [] - for k in _table: - if k not in _numchars: - _todelete.append(k) - _todelete = ''.join(_todelete) - del k - - def _eval(astr): - str_ = astr.translate(_table, _todelete) - if not str_: - raise TypeError("Invalid data string supplied: " + astr) - else: - return eval(str_) - def _convert_from_string(data): + for char in '[]': + data = data.replace(char, '') + rows = data.split(';') newdata = [] count = 0 @@ -54,7 +20,7 @@ def _convert_from_string(data): newrow = [] for col in trow: temp = col.split() - newrow.extend(map(_eval, temp)) + newrow.extend(map(ast.literal_eval, temp)) if count == 0: Ncols = len(newrow) elif len(newrow) != Ncols:

np.matrix('True False True') throws error When trying to create a numpy matrix using the string syntax `a = np.matrix('1 2; 3 4')` but with booleans instead of integers `np.matrix('True True False')` I see this error ``` >>> np.matrix('True True False') Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 267, in __new__ data = _convert_from_string(data) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 57, in _convert_from_string newrow.extend(map(_eval, temp)) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 26, in _eval return eval(str_) File "<string>", line 1, in <module> NameError: name 'e' is not defined ``` I'm not sure but is this desired behavior? Thanks!

2017-01-18T21:30:07Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 267, in __new__ data = _convert_from_string(data) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 57, in _convert_from_string newrow.extend(map(_eval, temp)) File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/numpy/matrixlib/defmatrix.py", line 26, in _eval return eval(str_) File "<string>", line 1, in <module> NameError: name 'e' is not defined

10,510

numpy/numpy

numpy__numpy-8508

a621a2b700415a5c155546f9cb1f064c6099579e

diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -6130,8 +6130,11 @@ def __new__(self): def __array_finalize__(self, obj): return - def __array_wrap__(self, obj): - return self + def __array_prepare__(self, obj, context=None): + return self.view(MaskedArray).__array_prepare__(obj, context) + + def __array_wrap__(self, obj, context=None): + return self.view(MaskedArray).__array_wrap__(obj, context) def __str__(self): return str(masked_print_option._display)

__array_prepare__ produces bad shape for np.ma.masked A somewhat contrived example ```python >>> source = np.ma.masked # comes from user function, could be anything >>> source = np.asanyarray(source) # force to array, so we can try to copy the type >>> outarr = np.zeros((2, 2)) # allocate a raw ndarray for the result >>> bad = source.__array_prepare__(outarr) # prepare that raw array for operations >>> bad.shape (2, 2) >>> bad.data.shape () # uh oh >>> bad.mask.shape () # spaghettios ``` Which leads to failures like ```python >>> res.transpose((1, 0)) Traceback (most recent call last): File "<pyshell#78>", line 1, in <module> np.asanyarray(np.ma.masked).__array_prepare__(np.zeros((2, 2))).transpose((1, 0)) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 2509, in wrapped_method result = getattr(self._data, funcname)(*args, **params) ValueError: axes don't match array ``` This come up when trying to invoke #8441 on masked arrays. Is this a bug, or an invalid use of `__array_prepare__`?

Related: ```python def assert_broadcasts(ufunc, a, b): expected = np.broadcast(a, b).shape actual = ufunc(a, b).shape assert expected == actual assert_broadcasts(np.add, [1, 2, 3], 1) assert_broadcasts(np.add, [1, 2, 3], np.ma.masked) # AssertionError ```

2017-01-20T18:16:20Z

[]

Traceback (most recent call last): File "<pyshell#78>", line 1, in <module> np.asanyarray(np.ma.masked).__array_prepare__(np.zeros((2, 2))).transpose((1, 0)) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 2509, in wrapped_method result = getattr(self._data, funcname)(*args, **params) ValueError: axes don't match array

10,511

numpy/numpy

numpy__numpy-8647

b097bd7ed4fa8e574414c2a0df8e50ac27bffa6d

diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -6991,44 +6991,42 @@ def where(condition, x=_NoValue, y=_NoValue): [6.0 -- 8.0]] """ - missing = (x is _NoValue, y is _NoValue).count(True) + # handle the single-argument case + missing = (x is _NoValue, y is _NoValue).count(True) if missing == 1: raise ValueError("Must provide both 'x' and 'y' or neither.") if missing == 2: - return filled(condition, 0).nonzero() - - # Both x and y are provided - - # Get the condition - fc = filled(condition, 0).astype(MaskType) - notfc = np.logical_not(fc) - - # Get the data - xv = getdata(x) - yv = getdata(y) - if x is masked: - ndtype = yv.dtype - elif y is masked: - ndtype = xv.dtype - else: - ndtype = np.find_common_type([xv.dtype, yv.dtype], []) - - # Construct an empty array and fill it - d = np.empty(fc.shape, dtype=ndtype).view(MaskedArray) - np.copyto(d._data, xv.astype(ndtype), where=fc) - np.copyto(d._data, yv.astype(ndtype), where=notfc) - - # Create an empty mask and fill it - mask = np.zeros(fc.shape, dtype=MaskType) - np.copyto(mask, getmask(x), where=fc) - np.copyto(mask, getmask(y), where=notfc) - mask |= getmaskarray(condition) - - # Use d._mask instead of d.mask to avoid copies - d._mask = mask if mask.any() else nomask + return nonzero(condition) + + # we only care if the condition is true - false or masked pick y + cf = filled(condition, False) + xd = getdata(x) + yd = getdata(y) + + # we need the full arrays here for correct final dimensions + cm = getmaskarray(condition) + xm = getmaskarray(x) + ym = getmaskarray(y) + + # deal with the fact that masked.dtype == float64, but we don't actually + # want to treat it as that. + if x is masked and y is not masked: + xd = np.zeros((), dtype=yd.dtype) + xm = np.ones((), dtype=ym.dtype) + elif y is masked and x is not masked: + yd = np.zeros((), dtype=xd.dtype) + ym = np.ones((), dtype=xm.dtype) + + data = np.where(cf, xd, yd) + mask = np.where(cf, xm, ym) + mask = np.where(cm, np.ones((), dtype=mask.dtype), mask) + + # collapse the mask, for backwards compatibility + if mask.dtype == np.bool_ and not mask.any(): + mask = nomask - return d + return masked_array(data, mask=mask) def choose(indices, choices, out=None, mode='raise'):

BUG: np.ma.where does not broadcast correctly ```python >>> x = np.eye(3) >>> y = np.eye(3) >>> np.where([0, 1, 0], x, y) array([[ 1., 0., 0.], [ 0., 1., 0.], [ 0., 0., 1.]]) >>> np.ma.where([0, 1, 0], x, y) Traceback (most recent call last): File "<pyshell#24>", line 1, in <module> np.ma.where([0, 1, 0], x, y) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 6964, in where np.copyto(d._data, xv.astype(ndtype), where=fc) ValueError: could not broadcast input array from shape (3,3) into shape (3) ``` BUG: np.ma.where does not handle structured dtypes correctly ```python >>> dt = np.dtype([('a', int), ('b', int)]) >>> x = np.array([(1, 2), (3, 4)], dtype=dt) >>> np.where([0, 1], x, np.array((2, 3), dtype=dt)).dtype dtype([('a', '<i4'), ('b', '<i4')]) >>> np.ma.where([0, 1], x, np.array((2, 3), dtype=dt)).dtype dtype('O') ```

2017-02-20T15:20:04Z

[]

Traceback (most recent call last): File "<pyshell#24>", line 1, in <module> np.ma.where([0, 1, 0], x, y) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\ma\core.py", line 6964, in where np.copyto(d._data, xv.astype(ndtype), where=fc) ValueError: could not broadcast input array from shape (3,3) into shape (3)

10,522

numpy/numpy

numpy__numpy-8665

85cd7b7e1ef04a498dbb84fc7d7fb35881a73183

diff --git a/numpy/ma/core.py b/numpy/ma/core.py --- a/numpy/ma/core.py +++ b/numpy/ma/core.py @@ -6372,21 +6372,16 @@ def getdoc(self): def __call__(self, a, *args, **params): if self.reversed: args = list(args) - arr = args[0] - args[0] = a - a = arr - # Get the method from the array (if possible) + a, args[0] = args[0], a + + marr = asanyarray(a) method_name = self.__name__ - method = getattr(a, method_name, None) - if method is not None: - return method(*args, **params) - # Still here ? Then a is not a MaskedArray - method = getattr(MaskedArray, method_name, None) - if method is not None: - return method(MaskedArray(a), *args, **params) - # Still here ? OK, let's call the corresponding np function - method = getattr(np, method_name) - return method(a, *args, **params) + method = getattr(type(marr), method_name, None) + if method is None: + # use the corresponding np function + method = getattr(np, method_name) + + return method(marr, *args, **params) all = _frommethod('all') @@ -6535,9 +6530,7 @@ def compressed(x): Equivalent method. """ - if not isinstance(x, MaskedArray): - x = asanyarray(x) - return x.compressed() + return asanyarray(x).compressed() def concatenate(arrays, axis=0): @@ -7683,6 +7676,10 @@ def asanyarray(a, dtype=None): <class 'numpy.ma.core.MaskedArray'> """ + # workaround for #8666, to preserve identity. Ideally the bottom line + # would handle this for us. + if isinstance(a, MaskedArray) and (dtype is None or dtype == a.dtype): + return a return masked_array(a, dtype=dtype, copy=False, keep_mask=True, subok=True)

np.ma.count and np.ma.copy on list input The `np.ma.count` and `np.ma.copy` behave strangly if the input is a `list` or `tuple` (probably also other python builtins). ``` >>> np.ma.copy([1,2,3]) # unexpected behaviour [1, 2, 3] >>> np.copy([1,2,3]) # expected behaviour array([1, 2, 3]) >>> np.ma.count([1,2,3]) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Programming\Anaconda\envs\test\lib\site-packages\numpy\ma\core.py", line 6389, in __call__ return method(*args, **params) TypeError: count() takes exactly one argument (0 given) ``` The reason for this behaviour is that `list` **has** a `count` and `copy` method and `np.ma._frommethod` tries to call it instead of `np.ma.MaskedArray.count` (and copy).

2017-02-22T10:22:33Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Programming\Anaconda\envs\test\lib\site-packages\numpy\ma\core.py", line 6389, in __call__ return method(*args, **params) TypeError: count() takes exactly one argument (0 given)

10,524

numpy/numpy

numpy__numpy-8750

6a3edf3210b439a4d1a51acb4e01bac017697ee6

diff --git a/numpy/lib/function_base.py b/numpy/lib/function_base.py --- a/numpy/lib/function_base.py +++ b/numpy/lib/function_base.py @@ -1135,7 +1135,7 @@ def average(a, axis=None, weights=None, returned=False): wgt = wgt.swapaxes(-1, axis) scl = wgt.sum(axis=axis, dtype=result_dtype) - if (scl == 0.0).any(): + if np.any(scl == 0.0): raise ZeroDivisionError( "Weights sum to zero, can't be normalized")

np.average crashes for 1D decimal object array (now only without weights) I'm running macOS 10.12.3, Python 3.6.0, and NumPy 1.12.0. When using the decimal package with NumPy, the arrays are of type 'object'. While most operations are performed flawlessly, calling np.average() with one of these arrays throws an error if weights are not provided (`AttributeError: 'decimal.Decimal' object has no attribute 'dtype'`), and a different error if weights are provided (`AttributeError: 'bool' object has no attribute 'any'`). Contrarily, calling np.mean() executes without error. The source of the problem in both cases appears to be the assumption that the result of folding the input array results in a standard supported dtype, rather than being based on the operations available to the object type. Therefore, I suspect that this issue will extend to any types which support the requisite numeric operations but are not natively supported by NumPy. I resolved the weighted problem for my use case by changing the line `if (scl == 0.0).any()` (line 1138 in `lib/function_base.py`) to check if scl were an array first, and if not then removing the `any()` call, though I don't know if that solution is desirable or acceptable for the purposes of NumPy. Sample input: ``` import numpy as np import decimal as dc values = np.array([dc.Decimal(x) for x in range(10)]) weights = np.array([dc.Decimal(x) for x in range(10)]) weights /= weights.sum() print(np.mean(values)) print(np.average(values, weights=weights)) ``` Corresponding output: ``` 4.5 Traceback (most recent call last): File "bugreport.py", line 9, in <module> print(np.average(values, weights=weights)) File "/usr/local/lib/python3.6/site-packages/numpy/lib/function_base.py", line 1138, in average if (scl == 0.0).any(): AttributeError: 'bool' object has no attribute 'any' ```

Yes, this seems like a bug and certainly doesn't reflected the intended behavior as described in the docs. Decimal arrays should be acceptable as arguments for this function.

2017-03-06T20:07:37Z

[]

Traceback (most recent call last): File "bugreport.py", line 9, in <module> print(np.average(values, weights=weights)) File "/usr/local/lib/python3.6/site-packages/numpy/lib/function_base.py", line 1138, in average if (scl == 0.0).any(): AttributeError: 'bool' object has no attribute 'any'

10,531

numpy/numpy

numpy__numpy-8762

485b099cd4b82d65dc38cb2b28c7119f003c76c4

diff --git a/numpy/lib/polynomial.py b/numpy/lib/polynomial.py --- a/numpy/lib/polynomial.py +++ b/numpy/lib/polynomial.py @@ -1036,17 +1036,47 @@ class poly1d(object): poly1d([ 1, -3, 2]) """ - coeffs = None - order = None - variable = None __hash__ = None - def __init__(self, c_or_r, r=0, variable=None): + @property + def coeffs(self): + """ The polynomial coefficients """ + return self._coeffs + + @property + def variable(self): + """ The name of the polynomial variable """ + return self._variable + + # calculated attributes + @property + def order(self): + """ The order or degree of the polynomial """ + return len(self._coeffs) - 1 + + @property + def roots(self): + """ The roots of the polynomial, where self(x) == 0 """ + return roots(self._coeffs) + + # alias attributes + r = roots + c = coef = coefficients = coeffs + o = order + + def __init__(self, c_or_r, r=False, variable=None): if isinstance(c_or_r, poly1d): - for key in c_or_r.__dict__.keys(): - self.__dict__[key] = c_or_r.__dict__[key] + self._variable = c_or_r._variable + self._coeffs = c_or_r._coeffs + + if set(c_or_r.__dict__) - set(self.__dict__): + msg = ("In the future extra properties will not be copied " + "across when constructing one poly1d from another") + warnings.warn(msg, FutureWarning, stacklevel=2) + self.__dict__.update(c_or_r.__dict__) + if variable is not None: - self.__dict__['variable'] = variable + self._variable = variable return if r: c_or_r = poly(c_or_r) @@ -1056,11 +1086,10 @@ def __init__(self, c_or_r, r=0, variable=None): c_or_r = trim_zeros(c_or_r, trim='f') if len(c_or_r) == 0: c_or_r = NX.array([0.]) - self.__dict__['coeffs'] = c_or_r - self.__dict__['order'] = len(c_or_r) - 1 + self._coeffs = c_or_r if variable is None: variable = 'x' - self.__dict__['variable'] = variable + self._variable = variable def __array__(self, t=None): if t: @@ -1199,29 +1228,17 @@ def __rdiv__(self, other): __rtruediv__ = __rdiv__ def __eq__(self, other): + if not isinstance(other, poly1d): + return NotImplemented if self.coeffs.shape != other.coeffs.shape: return False return (self.coeffs == other.coeffs).all() def __ne__(self, other): + if not isinstance(other, poly1d): + return NotImplemented return not self.__eq__(other) - def __setattr__(self, key, val): - raise ValueError("Attributes cannot be changed this way.") - - def __getattr__(self, key): - if key in ['r', 'roots']: - return roots(self.coeffs) - elif key in ['c', 'coef', 'coefficients']: - return self.coeffs - elif key in ['o']: - return self.order - else: - try: - return self.__dict__[key] - except KeyError: - raise AttributeError( - "'%s' has no attribute '%s'" % (self.__class__, key)) def __getitem__(self, val): ind = self.order - val @@ -1237,10 +1254,9 @@ def __setitem__(self, key, val): raise ValueError("Does not support negative powers.") if key > self.order: zr = NX.zeros(key-self.order, self.coeffs.dtype) - self.__dict__['coeffs'] = NX.concatenate((zr, self.coeffs)) - self.__dict__['order'] = key + self._coeffs = NX.concatenate((zr, self.coeffs)) ind = 0 - self.__dict__['coeffs'][ind] = val + self._coeffs[ind] = val return def __iter__(self):

numpy.poly1d.__eq__ method fails with AttributeError I called the `inspect.signature` function on a `numpy.poly1d` object: ```python import numpy as np import inspect poly = np.poly1d([1, 2, 3]) print(inspect.signature(poly)) ``` and got an unexpected exception: ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/python3.5/inspect.py", line 2987, in signature return Signature.from_callable(obj, follow_wrapped=follow_wrapped) File "/python3.5/inspect.py", line 2737, in from_callable follow_wrapper_chains=follow_wrapped) File "/python3.5/inspect.py", line 2228, in _signature_from_callable if _signature_is_builtin(obj): File "/python3.5/inspect.py", line 1785, in _signature_is_builtin obj in (type, object)) File "/python3.5/site-packages/numpy/lib/polynomial.py", line 1203, in __eq__ if self.coeffs.shape != other.coeffs.shape: AttributeError: type object 'type' has no attribute 'coeffs' ``` So `numpy.poly1d` objects can be compared only to objects with special attributes structure. The current`__eq__` implementation from [v.1.12.x](https://github.com/numpy/numpy/blob/maintenance/1.12.x/numpy/lib/polynomial.py#L1201-L1204): ```python def __eq__(self, other): if self.coeffs.shape != other.coeffs.shape: return False return (self.coeffs == other.coeffs).all() ``` numpy version 1.12.0

That should probably have a `if not isinstance(other, poly1d): return NotImplemented`. In fact, a whole bunch of the `__(.*)__` methods should have those lines I think `if not isinstance(other, poly1d): return False` is better. With `raise NotImplemented` we still will have the same problem. Is there any reason to not return False in such cases? > With raise NotImplemented we still will have the same problem. @bondarevts: Indeed, that is why I suggested `return NotImplemented` and not `raise NotImplementedError`. `NotImplemented` here just means "I have no idea, ask the other guy", and does not result in an error > Indeed, that is why I suggested return NotImplemented @eric-wieser You're right. I read it wrong. Thank you for the clarification.

2017-03-09T02:57:50Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/python3.5/inspect.py", line 2987, in signature return Signature.from_callable(obj, follow_wrapped=follow_wrapped) File "/python3.5/inspect.py", line 2737, in from_callable follow_wrapper_chains=follow_wrapped) File "/python3.5/inspect.py", line 2228, in _signature_from_callable if _signature_is_builtin(obj): File "/python3.5/inspect.py", line 1785, in _signature_is_builtin obj in (type, object)) File "/python3.5/site-packages/numpy/lib/polynomial.py", line 1203, in __eq__ if self.coeffs.shape != other.coeffs.shape: AttributeError: type object 'type' has no attribute 'coeffs'

10,532

numpy/numpy

numpy__numpy-8827

03f3789efe4da2c56d2841ed027ef6735ca2f11b

diff --git a/numpy/linalg/linalg.py b/numpy/linalg/linalg.py --- a/numpy/linalg/linalg.py +++ b/numpy/linalg/linalg.py @@ -19,12 +19,13 @@ import warnings from numpy.core import ( - array, asarray, zeros, empty, empty_like, transpose, intc, single, double, + array, asarray, zeros, empty, empty_like, intc, single, double, csingle, cdouble, inexact, complexfloating, newaxis, ravel, all, Inf, dot, add, multiply, sqrt, maximum, fastCopyAndTranspose, sum, isfinite, size, finfo, errstate, geterrobj, longdouble, moveaxis, amin, amax, product, abs, - broadcast, atleast_2d, intp, asanyarray, isscalar, object_, ones - ) + broadcast, atleast_2d, intp, asanyarray, isscalar, object_, ones, matmul, + swapaxes, divide) + from numpy.core.multiarray import normalize_axis_index from numpy.lib import triu, asfarray from numpy.linalg import lapack_lite, _umath_linalg @@ -223,6 +224,22 @@ def _assertNoEmpty2d(*arrays): if _isEmpty2d(a): raise LinAlgError("Arrays cannot be empty") +def transpose(a): + """ + Transpose each matrix in a stack of matrices. + + Unlike np.transpose, this only swaps the last two axes, rather than all of + them + + Parameters + ---------- + a : (...,M,N) array_like + + Returns + ------- + aT : (...,N,M) ndarray + """ + return swapaxes(a, -1, -2) # Linear equations @@ -1279,7 +1296,7 @@ def eigh(a, UPLO='L'): # Singular value decomposition -def svd(a, full_matrices=1, compute_uv=1): +def svd(a, full_matrices=True, compute_uv=True): """ Singular Value Decomposition. @@ -1494,15 +1511,21 @@ def matrix_rank(M, tol=None): Rank of the array is the number of SVD singular values of the array that are greater than `tol`. + .. versionchanged:: 1.14 + Can now operate on stacks of matrices + Parameters ---------- M : {(M,), (..., M, N)} array_like input vector or stack of matrices - tol : {None, float}, optional - threshold below which SVD values are considered zero. If `tol` is - None, and ``S`` is an array with singular values for `M`, and - ``eps`` is the epsilon value for datatype of ``S``, then `tol` is - set to ``S.max() * max(M.shape) * eps``. + tol : (...) array_like, float, optional + threshold below which SVD values are considered zero. If `tol` is + None, and ``S`` is an array with singular values for `M`, and + ``eps`` is the epsilon value for datatype of ``S``, then `tol` is + set to ``S.max() * max(M.shape) * eps``. + + .. versionchanged:: 1.14 + Broadcasted against the stack of matrices Notes ----- @@ -1569,6 +1592,8 @@ def matrix_rank(M, tol=None): S = svd(M, compute_uv=False) if tol is None: tol = S.max(axis=-1, keepdims=True) * max(M.shape[-2:]) * finfo(S.dtype).eps + else: + tol = asarray(tol)[...,newaxis] return (S > tol).sum(axis=-1) @@ -1582,26 +1607,29 @@ def pinv(a, rcond=1e-15 ): singular-value decomposition (SVD) and including all *large* singular values. + .. versionchanged:: 1.14 + Can now operate on stacks of matrices + Parameters ---------- - a : (M, N) array_like - Matrix to be pseudo-inverted. - rcond : float - Cutoff for small singular values. - Singular values smaller (in modulus) than - `rcond` * largest_singular_value (again, in modulus) - are set to zero. + a : (..., M, N) array_like + Matrix or stack of matrices to be pseudo-inverted. + rcond : (...) array_like of float + Cutoff for small singular values. + Singular values smaller (in modulus) than + `rcond` * largest_singular_value (again, in modulus) + are set to zero. Broadcasts against the stack of matrices Returns ------- - B : (N, M) ndarray - The pseudo-inverse of `a`. If `a` is a `matrix` instance, then so - is `B`. + B : (..., N, M) ndarray + The pseudo-inverse of `a`. If `a` is a `matrix` instance, then so + is `B`. Raises ------ LinAlgError - If the SVD computation does not converge. + If the SVD computation does not converge. Notes ----- @@ -1638,20 +1666,20 @@ def pinv(a, rcond=1e-15 ): """ a, wrap = _makearray(a) + rcond = asarray(rcond) if _isEmpty2d(a): res = empty(a.shape[:-2] + (a.shape[-1], a.shape[-2]), dtype=a.dtype) return wrap(res) a = a.conjugate() - u, s, vt = svd(a, 0) - m = u.shape[0] - n = vt.shape[1] - cutoff = rcond*maximum.reduce(s) - for i in range(min(n, m)): - if s[i] > cutoff: - s[i] = 1./s[i] - else: - s[i] = 0. - res = dot(transpose(vt), multiply(s[:, newaxis], transpose(u))) + u, s, vt = svd(a, full_matrices=False) + + # discard small singular values + cutoff = rcond[..., newaxis] * amax(s, axis=-1, keepdims=True) + large = s > cutoff + s = divide(1, s, where=large, out=s) + s[~large] = 0 + + res = matmul(transpose(vt), multiply(s[..., newaxis], transpose(u))) return wrap(res) # Determinant @@ -1987,13 +2015,13 @@ def lstsq(a, b, rcond="warn"): resids = array([sum((ravel(bstar)[n:])**2)], dtype=result_real_t) else: - x = array(transpose(bstar)[:n,:], dtype=result_t, copy=True) + x = array(bstar.T[:n,:], dtype=result_t, copy=True) if results['rank'] == n and m > n: if isComplexType(t): - resids = sum(abs(transpose(bstar)[n:,:])**2, axis=0).astype( + resids = sum(abs(bstar.T[n:,:])**2, axis=0).astype( result_real_t, copy=False) else: - resids = sum((transpose(bstar)[n:,:])**2, axis=0).astype( + resids = sum((bstar.T[n:,:])**2, axis=0).astype( result_real_t, copy=False) st = s[:min(n, m)].astype(result_real_t, copy=True)

BUG: Linalg.pinv fails on stacks of matrices ```python >>> a = np.stack((np.eye(3),)*4, axis=0) >>> ai = np.linalg.inv(a) >>> assert (a == ai).all() >>> api = np.linalg.pinv(a) Traceback (most recent call last): File "<pyshell#9>", line 1, in <module> np.linalg.pinv(a) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\linalg\linalg.py", line 1668, in pinv if s[i] > cutoff: ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() ```

2017-03-25T00:01:50Z

[]

Traceback (most recent call last): File "<pyshell#9>", line 1, in <module> np.linalg.pinv(a) File "C:\Program Files\Python 3.5\lib\site-packages\numpy\linalg\linalg.py", line 1668, in pinv if s[i] > cutoff: ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

10,536

numpy/numpy

numpy__numpy-9285

c6533b6c386dc0f4009e5f3c5c545dde4d1b48a4

diff --git a/numpy/distutils/ccompiler.py b/numpy/distutils/ccompiler.py --- a/numpy/distutils/ccompiler.py +++ b/numpy/distutils/ccompiler.py @@ -80,6 +80,7 @@ def _needs_build(obj, cc_args, extra_postargs, pp_opts): return False + def replace_method(klass, method_name, func): if sys.version_info[0] < 3: m = types.MethodType(func, None, klass) @@ -88,6 +89,25 @@ def replace_method(klass, method_name, func): m = lambda self, *args, **kw: func(self, *args, **kw) setattr(klass, method_name, m) + +###################################################################### +## Method that subclasses may redefine. But don't call this method, +## it i private to CCompiler class and may return unexpected +## results if used elsewhere. So, you have been warned.. + +def CCompiler_find_executables(self): + """ + Does nothing here, but is called by the get_version method and can be + overridden by subclasses. In particular it is redefined in the `FCompiler` + class where more documentation can be found. + + """ + pass + + +replace_method(CCompiler, 'find_executables', CCompiler_find_executables) + + # Using customized CCompiler.spawn. def CCompiler_spawn(self, cmd, display=None): """ diff --git a/numpy/distutils/fcompiler/intel.py b/numpy/distutils/fcompiler/intel.py --- a/numpy/distutils/fcompiler/intel.py +++ b/numpy/distutils/fcompiler/intel.py @@ -57,7 +57,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-xhost -fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): @@ -123,7 +123,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): diff --git a/numpy/distutils/intelccompiler.py b/numpy/distutils/intelccompiler.py --- a/numpy/distutils/intelccompiler.py +++ b/numpy/distutils/intelccompiler.py @@ -19,7 +19,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe @@ -59,7 +59,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -m64 -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe

numpy 1.13.0 doesn't build with Intel compilers I'm running into the problem below when trying to build `numpy` 1.13.0 with Intel compilers; building earlier numpy versions (1.12.1, 1.11.1, 1.10.4, ...) with this approach works fine. ``` $ python setup.py build --compiler=intel --fcompiler=intelem building library "npymath" sources Found executable /path/to/icc Could not locate executable ecc Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(**metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(**new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables' ``` The `Could not locate executable ecc` seems to be a hint, since this message doesn't occur with older numpy versions where building like this does work: ``` ... building library "npymath" sources Found executable /path/to/icc customize IntelEM64TFCompiler Found executable /path/to/ifort customize IntelEM64TFCompiler using config C compiler: icc -fPIC -fp-model strict -O3 -fomit-frame-pointer -openmp ... ``` Why is `numpy` looking for `ecc` when it found `icc`, and not continuing to go and find `ifort` like it did before? A similar problem was reported by @Eric89GXL in #9101, but there the problem was that `icc` could not be found, that's clearly not the problem here. Seeing this with Python 2.7.13, Intel compilers 2017.1.132, Intel MKL 2017.1.132.

I get the same error. Commenting out the line `self.find_executables()` works as a work-around. This method seems to be undefined for CCompiler/IntelCCompiler. #8961 is the only change from 1.12.1 and I don't see how that would be related to this. #8961 has added exactly the line causing this (`v = self.get_version()`). So an alternate fix to #8961 suggested in #8941 was `-fopenmp`, which would not be version dependent. Might want to try that. Hi everybody, In regard to this change, in numpy 1.13.0 I've found using f2py and intel compilers I got the following compilation error: `mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' AttributeError: LooseVersion instance has no attribute 'split'` which can be solved changing the above line by: `mpopt = 'openmp' if v and int(v.version[0]) < 15 else 'qopenmp'` Hope this info can be useful, ACM Geez, version is a `LooseVersion` instance. Seems kinda useless given the lack of documentation of the class and the variations of version in the wild. That could be fixed by calling `str` on it I suppose. So there are two problems: * `LooseVersion` is not a string * missing `find_executables`, which only seems to be defined for fortran compilers @boegel Looks like "intel" is so generic that the Itanium platform is also being searched, weird. EDIT: I don't think the ecc message is relevant here.

2017-06-21T15:59:33Z

[]

Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(**metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(**new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables'

10,556

numpy/numpy

numpy__numpy-9299

753e148e424e89da60390d7f1e15e65153da5aa0

diff --git a/numpy/distutils/ccompiler.py b/numpy/distutils/ccompiler.py --- a/numpy/distutils/ccompiler.py +++ b/numpy/distutils/ccompiler.py @@ -80,6 +80,7 @@ def _needs_build(obj, cc_args, extra_postargs, pp_opts): return False + def replace_method(klass, method_name, func): if sys.version_info[0] < 3: m = types.MethodType(func, None, klass) @@ -88,6 +89,25 @@ def replace_method(klass, method_name, func): m = lambda self, *args, **kw: func(self, *args, **kw) setattr(klass, method_name, m) + +###################################################################### +## Method that subclasses may redefine. But don't call this method, +## it i private to CCompiler class and may return unexpected +## results if used elsewhere. So, you have been warned.. + +def CCompiler_find_executables(self): + """ + Does nothing here, but is called by the get_version method and can be + overridden by subclasses. In particular it is redefined in the `FCompiler` + class where more documentation can be found. + + """ + pass + + +replace_method(CCompiler, 'find_executables', CCompiler_find_executables) + + # Using customized CCompiler.spawn. def CCompiler_spawn(self, cmd, display=None): """ diff --git a/numpy/distutils/fcompiler/intel.py b/numpy/distutils/fcompiler/intel.py --- a/numpy/distutils/fcompiler/intel.py +++ b/numpy/distutils/fcompiler/intel.py @@ -57,7 +57,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-xhost -fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): @@ -123,7 +123,7 @@ def get_flags(self): def get_flags_opt(self): # Scipy test failures with -O2 v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' return ['-fp-model strict -O1 -{}'.format(mpopt)] def get_flags_arch(self): diff --git a/numpy/distutils/intelccompiler.py b/numpy/distutils/intelccompiler.py --- a/numpy/distutils/intelccompiler.py +++ b/numpy/distutils/intelccompiler.py @@ -19,7 +19,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe @@ -59,7 +59,7 @@ def __init__(self, verbose=0, dry_run=0, force=0): UnixCCompiler.__init__(self, verbose, dry_run, force) v = self.get_version() - mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' + mpopt = 'openmp' if v and v < '15' else 'qopenmp' self.cc_exe = ('icc -m64 -fPIC -fp-model strict -O3 ' '-fomit-frame-pointer -{}').format(mpopt) compiler = self.cc_exe

numpy 1.13.0 doesn't build with Intel compilers I'm running into the problem below when trying to build `numpy` 1.13.0 with Intel compilers; building earlier numpy versions (1.12.1, 1.11.1, 1.10.4, ...) with this approach works fine. ``` $ python setup.py build --compiler=intel --fcompiler=intelem building library "npymath" sources Found executable /path/to/icc Could not locate executable ecc Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(**metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(**new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables' ``` The `Could not locate executable ecc` seems to be a hint, since this message doesn't occur with older numpy versions where building like this does work: ``` ... building library "npymath" sources Found executable /path/to/icc customize IntelEM64TFCompiler Found executable /path/to/ifort customize IntelEM64TFCompiler using config C compiler: icc -fPIC -fp-model strict -O3 -fomit-frame-pointer -openmp ... ``` Why is `numpy` looking for `ecc` when it found `icc`, and not continuing to go and find `ifort` like it did before? A similar problem was reported by @Eric89GXL in #9101, but there the problem was that `icc` could not be found, that's clearly not the problem here. Seeing this with Python 2.7.13, Intel compilers 2017.1.132, Intel MKL 2017.1.132.

I get the same error. Commenting out the line `self.find_executables()` works as a work-around. This method seems to be undefined for CCompiler/IntelCCompiler. #8961 is the only change from 1.12.1 and I don't see how that would be related to this. #8961 has added exactly the line causing this (`v = self.get_version()`). So an alternate fix to #8961 suggested in #8941 was `-fopenmp`, which would not be version dependent. Might want to try that. Hi everybody, In regard to this change, in numpy 1.13.0 I've found using f2py and intel compilers I got the following compilation error: `mpopt = 'openmp' if v and int(v.split('.')[0]) < 15 else 'qopenmp' AttributeError: LooseVersion instance has no attribute 'split'` which can be solved changing the above line by: `mpopt = 'openmp' if v and int(v.version[0]) < 15 else 'qopenmp'` Hope this info can be useful, ACM Geez, version is a `LooseVersion` instance. Seems kinda useless given the lack of documentation of the class and the variations of version in the wild. That could be fixed by calling `str` on it I suppose. So there are two problems: * `LooseVersion` is not a string * missing `find_executables`, which only seems to be defined for fortran compilers @boegel Looks like "intel" is so generic that the Itanium platform is also being searched, weird. EDIT: I don't think the ecc message is relevant here.

2017-06-26T17:23:43Z

[]

Traceback (most recent call last): File "setup.py", line 392, in <module> setup_package() File "setup.py", line 384, in setup_package setup(**metadata) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/core.py", line 169, in setup return old_setup(**new_attr) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/core.py", line 151, in setup dist.run_commands() File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 953, in run_commands self.run_command(cmd) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build.py", line 47, in run old_build.run(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/build.py", line 127, in run self.run_command(cmd_name) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/cmd.py", line 326, in run_command self.distribution.run_command(command) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/dist.py", line 972, in run_command cmd_obj.run() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 148, in run self.build_sources() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 159, in build_sources self.build_library_sources(*libname_info) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 294, in build_library_sources sources = self.generate_sources(sources, (lib_name, build_info)) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/build_src.py", line 377, in generate_sources source = func(extension, build_dir) File "numpy/core/setup.py", line 672, in get_mathlib_info st = config_cmd.try_link('int main(void) { return 0;}') File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 248, in try_link self._check_compiler() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/command/config.py", line 39, in _check_compiler old_config._check_compiler(self) File "/user/home/gent/vsc400/vsc40023/eb_phanpyscratch/CO7/haswell-ib/software/Python/2.7.13-intel-2017a/lib/python2.7/distutils/command/config.py", line 102, in _check_compiler dry_run=self.dry_run, force=1) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 713, in new_compiler compiler = klass(None, dry_run, force) File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/intelccompiler.py", line 21, in __init__ v = self.get_version() File "/tmp/vsc40023/easybuild_build/numpy/1.13.0/intel-2017a-Python-2.7.13/numpy-1.13.0/numpy/distutils/ccompiler.py", line 582, in CCompiler_get_version self.find_executables() AttributeError: IntelCCompiler instance has no attribute 'find_executables'

10,557

numpy/numpy

numpy__numpy-9552

ae17d2c93dfac88cca9859d8b49490deb3991f41

diff --git a/numpy/doc/basics.py b/numpy/doc/basics.py --- a/numpy/doc/basics.py +++ b/numpy/doc/basics.py @@ -158,8 +158,8 @@ numpy provides with ``np.finfo(np.longdouble)``. NumPy does not provide a dtype with more precision than C -``long double``\s; in particular, the 128-bit IEEE quad precision -data type (FORTRAN's ``REAL*16``\) is not available. +``long double``\\s; in particular, the 128-bit IEEE quad precision +data type (FORTRAN's ``REAL*16``\\) is not available. For efficient memory alignment, ``np.longdouble`` is usually stored padded with zero bits, either to 96 or 128 bits. Which is more efficient

test_warning_calls "invalid escape sequence \s" on Python 3.6 daily wheel builds. New errors today (13 August), for Linux and OSX: ``` ====================================================================== ERROR: numpy.tests.test_warnings.test_warning_calls ---------------------------------------------------------------------- Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 79, in test_warning_calls tree = ast.parse(file.read()) File "/usr/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 184 SyntaxError: invalid escape sequence \s ``` https://travis-ci.org/MacPython/numpy-wheels/jobs/264023630 https://travis-ci.org/MacPython/numpy-wheels/jobs/264023631 https://travis-ci.org/MacPython/numpy-wheels/jobs/264023635

Was worried about that, the problem is ``` file: numpy/doc/basics.py line: 184 : invalid escape sequence \s ``` Came from documentation fixes. Probably need a raw string, although I would prefer making the files genuine rst. I should add that check to the tests...

2017-08-13T23:06:02Z

[]

Traceback (most recent call last): File "/venv/lib/python3.6/site-packages/nose/case.py", line 198, in runTest self.test(*self.arg) File "/venv/lib/python3.6/site-packages/numpy/tests/test_warnings.py", line 79, in test_warning_calls tree = ast.parse(file.read()) File "/usr/lib/python3.6/ast.py", line 35, in parse return compile(source, filename, mode, PyCF_ONLY_AST) File "<unknown>", line 184 SyntaxError: invalid escape sequence \s

10,567

open-mmlab/mmdetection

open-mmlab__mmdetection-10568

8822264e185df57250ac15bdbb86ac5a383e6520

diff --git a/demo/video_gpuaccel_demo.py b/demo/video_gpuaccel_demo.py --- a/demo/video_gpuaccel_demo.py +++ b/demo/video_gpuaccel_demo.py @@ -52,7 +52,9 @@ def prefetch_batch_input_shape(model: nn.Module, ori_wh: Tuple[int, test_pipeline = Compose(cfg.test_dataloader.dataset.pipeline) data = {'img': np.zeros((h, w, 3), dtype=np.uint8), 'img_id': 0} data = test_pipeline(data) - _, data_sample = model.data_preprocessor([data], False) + data['inputs'] = [data['inputs']] + data['data_samples'] = [data['data_samples']] + data_sample = model.data_preprocessor(data, False)['data_samples'] batch_input_shape = data_sample[0].batch_input_shape return batch_input_shape @@ -69,8 +71,8 @@ def pack_data(frame_resize: np.ndarray, batch_input_shape: Tuple[int, int], 'scale_factor': (batch_input_shape[0] / ori_shape[0], batch_input_shape[1] / ori_shape[1]) }) - frame_resize = torch.from_numpy(frame_resize).permute((2, 0, 1)) - data = {'inputs': frame_resize, 'data_sample': data_sample} + frame_resize = torch.from_numpy(frame_resize).permute((2, 0, 1)).cuda() + data = {'inputs': [frame_resize], 'data_samples': [data_sample]} return data @@ -112,7 +114,7 @@ def main(): for i, (frame_resize, frame_origin) in enumerate( zip(track_iter_progress(video_resize), video_origin)): data = pack_data(frame_resize, batch_input_shape, ori_shape) - result = model.test_step([data])[0] + result = model.test_step(data)[0] visualizer.add_datasample( name='video',

TypeError: list indices must be integers or slices, not str When I run the demo code **video_gpuaccel_demo.py**, it has the following error. How to solve it, thanks. Traceback (most recent call last): File "demo/video_gpuaccel_demo.py", line 147, in <module> main() File "demo/video_gpuaccel_demo.py", line 102, in main batch_input_shape = prefetch_batch_input_shape( File "demo/video_gpuaccel_demo.py", line 60, in prefetch_batch_input_shape _, data_sample = model.data_preprocessor([data], False) File "C:\Anaconda\Anaconda\envs\mmdetection\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 121, in forward batch_pad_shape = self._get_pad_shape(data) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 154, in _get_pad_shape _batch_inputs = data['inputs'] TypeError: list indices must be integers or slices, not str

2023-06-28T22:44:41Z

[]

Traceback (most recent call last): File "demo/video_gpuaccel_demo.py", line 147, in <module> main() File "demo/video_gpuaccel_demo.py", line 102, in main batch_input_shape = prefetch_batch_input_shape( File "demo/video_gpuaccel_demo.py", line 60, in prefetch_batch_input_shape _, data_sample = model.data_preprocessor([data], False) File "C:\Anaconda\Anaconda\envs\mmdetection\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 121, in forward batch_pad_shape = self._get_pad_shape(data) File "c:\research\programmar\deeplearning\vehicle_classification\mmdet\models\data_preprocessors\data_preprocessor.py", line 154, in _get_pad_shape _batch_inputs = data['inputs'] TypeError: list indices must be integers or slices, not str

10,594

open-mmlab/mmdetection

open-mmlab__mmdetection-1099

b6712d4a9abe261b34b6a62f89ed3ed1fb88fae1

diff --git a/mmdet/core/bbox/__init__.py b/mmdet/core/bbox/__init__.py --- a/mmdet/core/bbox/__init__.py +++ b/mmdet/core/bbox/__init__.py @@ -1,4 +1,3 @@ -from .assign_sampling import assign_and_sample, build_assigner, build_sampler from .assigners import AssignResult, BaseAssigner, MaxIoUAssigner from .bbox_target import bbox_target from .geometry import bbox_overlaps @@ -9,6 +8,9 @@ bbox_mapping, bbox_mapping_back, delta2bbox, distance2bbox, roi2bbox) +from .assign_sampling import ( # isort:skip, avoid recursive imports + assign_and_sample, build_assigner, build_sampler) + __all__ = [ 'bbox_overlaps', 'BaseAssigner', 'MaxIoUAssigner', 'AssignResult', 'BaseSampler', 'PseudoSampler', 'RandomSampler',

ImportError: cannot import name 'build_sampler' from 'mmdet.core.bbox.assign_sampling' I have successful install the mmdetection by the command "pip install -v -e .". But I have the problem in the test. Would anyone help me ? (lab) gpuserver@ubuntu:~/ht/labs/mmdetection-master$ python Python 3.7.3 (default, Mar 27 2019, 22:11:17) [GCC 7.3.0] :: Anaconda, Inc. on linux Type "help", "copyright", "credits" or "license" for more information. >>> from mmdet.apis import init_detector Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/__init__.py", line 2, in <module> from .inference import inference_detector, init_detector, show_result File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/inference.py", line 9, in <module> from mmdet.core import get_classes File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/__init__.py", line 1, in <module> from .anchor import * # noqa: F401, F403 File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/__init__.py", line 2, in <module> from .anchor_target import anchor_inside_flags, anchor_target File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/anchor_target.py", line 3, in <module> from ..bbox import PseudoSampler, assign_and_sample, bbox2delta, build_assigner File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/__init__.py", line 1, in <module> from .assign_sampling import assign_and_sample, build_assigner, build_sampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py", line 3, in <module> from . import assigners, samplers File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/__init__.py", line 2, in <module> from .combined_sampler import CombinedSampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/combined_sampler.py", line 1, in <module> from ..assign_sampling import build_sampler ImportError: cannot import name 'build_sampler' from 'mmdet.core.bbox.assign_sampling' (/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py)

I encountered the same error today. same problem here.

2019-08-01T09:48:27Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/__init__.py", line 2, in <module> from .inference import inference_detector, init_detector, show_result File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/apis/inference.py", line 9, in <module> from mmdet.core import get_classes File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/__init__.py", line 1, in <module> from .anchor import * # noqa: F401, F403 File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/__init__.py", line 2, in <module> from .anchor_target import anchor_inside_flags, anchor_target File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/anchor/anchor_target.py", line 3, in <module> from ..bbox import PseudoSampler, assign_and_sample, bbox2delta, build_assigner File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/__init__.py", line 1, in <module> from .assign_sampling import assign_and_sample, build_assigner, build_sampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py", line 3, in <module> from . import assigners, samplers File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/__init__.py", line 2, in <module> from .combined_sampler import CombinedSampler File "/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/samplers/combined_sampler.py", line 1, in <module> from ..assign_sampling import build_sampler ImportError: cannot import name 'build_sampler' from 'mmdet.core.bbox.assign_sampling' (/home/gpuserver/ht/labs/mmdetection-master/mmdet/core/bbox/assign_sampling.py)

10,596

open-mmlab/mmdetection

open-mmlab__mmdetection-1404

c64beaf1494ea68562b274a723824d7f52fd31e8

diff --git a/mmdet/datasets/loader/sampler.py b/mmdet/datasets/loader/sampler.py --- a/mmdet/datasets/loader/sampler.py +++ b/mmdet/datasets/loader/sampler.py @@ -132,8 +132,12 @@ def __iter__(self): math.ceil( size * 1.0 / self.samples_per_gpu / self.num_replicas) ) * self.samples_per_gpu * self.num_replicas - len(indice) - indice += indice[:extra] - indices += indice + # pad indice + tmp = indice.copy() + for _ in range(extra // size): + indice.extend(tmp) + indice.extend(tmp[:extra % size]) + indices.extend(indice) assert len(indices) == self.total_size

assert len(indices) == self.total_size error during multiple GPU training I am trying to train my dataset on 8 GPU's. However, after calling `./dist_train.sh` this error assertion appeares: Traceback (most recent call last): File "./tools/train.py", line 113, in <module> main() File "./tools/train.py", line 109, in main logger=logger) File "/mmdetection/mmdet/apis/train.py", line 58, in train_detector _dist_train(model, dataset, cfg, validate=validate) File "/mmdetection/mmdet/apis/train.py", line 186, in _dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 358, in run epoch_runner(data_loaders[i], **kwargs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 260, in train for i, data_batch in enumerate(data_loader): File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 193, in __iter__ return _DataLoaderIter(self) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 493, in __init__ self._put_indices() File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 591, in _put_indices indices = next(self.sample_iter, None) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 172, in __iter__ for idx in self.sampler: File "/mmdetection/mmdet/datasets/loader/sampler.py", line 138, in __iter__ assert len(indices) == self.total_size ... in the config I tried various values for `imgs_per_gpu` and `workers_per_gpu`, currently it is: `imgs_per_gpu=2, workers_per_gpu=2,` no settings was working though. Single-GPU training works well. What is the meaning of this assert? Thanks!

Please follow the `Error report` issue template. > Please follow the `Error report` issue template. Here it is, thanks for any help! **Checklist** 1. I have searched related issues but cannot get the expected help. yes 2. The bug has not been fixed in the latest version. yup **Describe the bug** I am trying to train my custom dataset on 8 GPU's. However, after calling ./dist_train.sh the error showed below appeares. In the config I tried more values for `imgs_per_gpu` and `workers_per_gpu` (e.g. `imgs_per_gpu=2`, `workers_per_gpu=2`), no settings was working though. Single-GPU training works well. What is the meaning of the assert in the Traceback? What does not fit? Thanks! **Reproduction** 1. What command or script did you run? ``` ./tools/dist_train.sh MY_CONFIG 8 --validate ``` 2. Did you make any modifications on the code or config? Did you understand what you have modified? I modified number of classes, workers_per_gpu, imgs_per_gpu, dataset type and paths to the datasets. No changes in code. 3. What dataset did you use? My own dataset of 8 classes converted to COCO format. **Environment** - OS: Ubuntu 18.04 - GCC 5.4.0 - PyTorch version 1.1.0 - I built the docker I use on the official pytorch+cuda docker - GPU model: 8xV100 - CUDA version: 10.0, CUDNN version: 7.5 **Error traceback** ``` Traceback (most recent call last): File "./tools/train.py", line 113, in main() File "./tools/train.py", line 109, in main logger=logger) File "/mmdetection/mmdet/apis/train.py", line 58, in train_detector _dist_train(model, dataset, cfg, validate=validate) File "/mmdetection/mmdet/apis/train.py", line 186, in _dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 358, in run epoch_runner(data_loaders[i], **kwargs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 260, in train for i, data_batch in enumerate(data_loader): File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 193, in iter return _DataLoaderIter(self) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 493, in init self._put_indices() File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 591, in _put_indices indices = next(self.sample_iter, None) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 172, in iter for idx in self.sampler: File "/mmdetection/mmdet/datasets/loader/sampler.py", line 138, in iter assert len(indices) == self.total_size ``` One more detail, when I print `len(indices)` and `self.total_size` right before the critical assert, it's `9308` and `9312`. The size of my training dataset is `9306`.. @hellock Any ideas? Recently I found out that when I set the config file to train with 2 GPUs (2 img/gpu), the training (called with `./dist_train`) initiates well. However, training with any more GPUs results in the mentioned assert error. Seems like a bug to me. You may count the images with aspect ratio >1 and <1. I suspect that there are only 2 images for one of the two groups. > You may count the images with aspect ratio >1 and <1. I suspect that there are only 2 images for one of the two groups. you are right, there are exactly 2 images with height > width The problem lies in [this line](https://github.com/open-mmlab/mmdetection/blob/master/mmdet/datasets/loader/sampler.py#L135) where `len(indices) < extra`. I meet the same issue, how to fix it? @ZhexuanZhou Before bug fixed, simplest way is make your img_per_gpu as power of 2. e.g. 2, 4, 8, 16 ... Also, make your gpu number as power of 2. This works for me. > @ZhexuanZhou > Before bug fixed, simplest way is make your img_per_gpu as power of 2. > e.g. 2, 4, 8, 16 ... > Also, make your gpu number as power of 2. > This works for me. This fix doesn't work for me, as is already mentioned in the bug description > **Describe the bug** > I am trying to train my custom dataset on 8 GPU's. However, after calling ./dist_train.sh the error showed below appeares. In the config I tried more values for `imgs_per_gpu` and `workers_per_gpu` (e.g. `imgs_per_gpu=2`, `workers_per_gpu=2`), no settings was working though. The easiest work-around for me was to comment out the two asserts in the `sampler.py` :) @FilipLangr Yeah, that might cause some samples lost but not that harmful. Hello! I have situation, when 6 pics, where w>h. 33994 pics, when h>w. And 2 pics, where h==w. I have deleted h==w pics and have AssertionError: assert len(indices) == self.total_size anyway :( Than I deleted w>h pics and get another error: TypeError: 'NoneType' object is not subscriptable

2019-09-16T10:01:55Z

[]

Traceback (most recent call last): File "./tools/train.py", line 113, in <module> main() File "./tools/train.py", line 109, in main logger=logger) File "/mmdetection/mmdet/apis/train.py", line 58, in train_detector _dist_train(model, dataset, cfg, validate=validate) File "/mmdetection/mmdet/apis/train.py", line 186, in _dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 358, in run epoch_runner(data_loaders[i], **kwargs) File "/opt/conda/lib/python3.6/site-packages/mmcv/runner/runner.py", line 260, in train for i, data_batch in enumerate(data_loader): File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 193, in __iter__ return _DataLoaderIter(self) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 493, in __init__ self._put_indices() File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 591, in _put_indices indices = next(self.sample_iter, None) File "/opt/conda/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 172, in __iter__ for idx in self.sampler: File "/mmdetection/mmdet/datasets/loader/sampler.py", line 138, in __iter__ assert len(indices) == self.total_size ... in the config I tried various values for `imgs_per_gpu` and `workers_per_gpu`, currently it is:

10,598

open-mmlab/mmdetection

open-mmlab__mmdetection-2030

aade6801e7df66679b1fe9d162da0d03b4742dd4

diff --git a/mmdet/datasets/pipelines/transforms.py b/mmdet/datasets/pipelines/transforms.py --- a/mmdet/datasets/pipelines/transforms.py +++ b/mmdet/datasets/pipelines/transforms.py @@ -156,7 +156,7 @@ def _resize_masks(self, results): mmcv.imresize(mask, mask_size, interpolation='nearest') for mask in results[key] ] - results[key] = masks + results[key] = np.stack(masks) def _resize_seg(self, results): for key in results.get('seg_fields', []): @@ -245,10 +245,10 @@ def __call__(self, results): results['flip_direction']) # flip masks for key in results.get('mask_fields', []): - results[key] = [ + results[key] = np.stack([ mmcv.imflip(mask, direction=results['flip_direction']) for mask in results[key] - ] + ]) # flip segs for key in results.get('seg_fields', []): @@ -410,7 +410,7 @@ def __call__(self, results): gt_mask = results['gt_masks'][i][crop_y1:crop_y2, crop_x1:crop_x2] valid_gt_masks.append(gt_mask) - results['gt_masks'] = valid_gt_masks + results['gt_masks'] = np.stack(valid_gt_masks) return results @@ -586,7 +586,7 @@ def __call__(self, results): 0).astype(mask.dtype) expand_mask[top:top + h, left:left + w] = mask expand_gt_masks.append(expand_mask) - results['gt_masks'] = expand_gt_masks + results['gt_masks'] = np.stack(expand_gt_masks) # not tested if 'gt_semantic_seg' in results: @@ -678,10 +678,10 @@ def __call__(self, results): results['gt_masks'][i] for i in range(len(mask)) if mask[i] ] - results['gt_masks'] = [ + results['gt_masks'] = np.stack([ gt_mask[patch[1]:patch[3], patch[0]:patch[2]] for gt_mask in valid_masks - ] + ]) # not tested if 'gt_semantic_seg' in results:

Stacking of masks done in `Pad` Hey guys, I run into this error when commenting out the line ``` dict(type='Pad', size_divisor=32), ``` in `train_pipeline` on custom images. It causes this error: ``` Traceback (most recent call last): File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main "__main__", mod_spec) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/__main__.py", line 45, in <module> cli.main() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 361, in main run() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 203, in run_file runpy.run_path(options.target, run_name="__main__") File "/usr/lib/python3.6/runpy.py", line 263, in run_path pkg_name=pkg_name, script_name=fname) File "/usr/lib/python3.6/runpy.py", line 96, in _run_module_code mod_name, mod_spec, pkg_name, script_name) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/dev/mmdetection/tools/train.py", line 124, in <module> main() File "/home/pc/dev/mmdetection/tools/train.py", line 120, in main timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 133, in train_detector timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 319, in _non_dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 364, in run epoch_runner(data_loaders[i], **kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 268, in train self.model, data_batch, train_mode=True, **kwargs) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 100, in batch_processor losses = model(**data) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 150, in forward return self.module(*inputs[0], **kwargs[0]) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "/home/pc/dev/mmdetection/mmdet/core/fp16/decorators.py", line 49, in new_func return old_func(*args, **kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/base.py", line 138, in forward return self.forward_train(img, img_meta, **kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/two_stage.py", line 254, in forward_train sampling_results, gt_masks, self.train_cfg.rcnn) File "/home/pc/dev/mmdetection/mmdet/models/mask_heads/fcn_mask_head.py", line 111, in get_target gt_masks, rcnn_train_cfg) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 12, in mask_target mask_targets = torch.cat(list(mask_targets)) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 22, in mask_target_single _, maxh, maxw = gt_masks.shape AttributeError: 'list' object has no attribute 'shape' ``` I already debugged it and as far as I have found out it is because the masks are stacked in [`_pad_masks`](https://github.com/open-mmlab/mmdetection/blob/10c82efb0392fc1a5e1c696a53fe9ca7dfc3cdda/mmdet/datasets/pipelines/transforms.py#L304) If padding is not included in the pipeline, however, then `gt_masks` is a `list` instead of `ndarray` causing the above error. Shouldn't the stacking be done in [`_load_masks`](https://github.com/open-mmlab/mmdetection/blob/10c82efb0392fc1a5e1c696a53fe9ca7dfc3cdda/mmdet/datasets/pipelines/loading.py#L82) already to have a more flexible pipeline?

Thanks for the suggestions. It is a known issue and we are preparing a PR to fix it. Ok, thank you!

2020-01-31T04:36:49Z

[]

Traceback (most recent call last): File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main "__main__", mod_spec) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/__main__.py", line 45, in <module> cli.main() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 361, in main run() File "/home/pc/.vscode-server/extensions/ms-python.python-2020.1.58038/pythonFiles/lib/python/new_ptvsd/no_wheels/ptvsd/../ptvsd/server/cli.py", line 203, in run_file runpy.run_path(options.target, run_name="__main__") File "/usr/lib/python3.6/runpy.py", line 263, in run_path pkg_name=pkg_name, script_name=fname) File "/usr/lib/python3.6/runpy.py", line 96, in _run_module_code mod_name, mod_spec, pkg_name, script_name) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pc/dev/mmdetection/tools/train.py", line 124, in <module> main() File "/home/pc/dev/mmdetection/tools/train.py", line 120, in main timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 133, in train_detector timestamp=timestamp) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 319, in _non_dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 364, in run epoch_runner(data_loaders[i], **kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/mmcv/runner/runner.py", line 268, in train self.model, data_batch, train_mode=True, **kwargs) File "/home/pc/dev/mmdetection/mmdet/apis/train.py", line 100, in batch_processor losses = model(**data) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 150, in forward return self.module(*inputs[0], **kwargs[0]) File "/home/pc/dev/venvs/mmdetection/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "/home/pc/dev/mmdetection/mmdet/core/fp16/decorators.py", line 49, in new_func return old_func(*args, **kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/base.py", line 138, in forward return self.forward_train(img, img_meta, **kwargs) File "/home/pc/dev/mmdetection/mmdet/models/detectors/two_stage.py", line 254, in forward_train sampling_results, gt_masks, self.train_cfg.rcnn) File "/home/pc/dev/mmdetection/mmdet/models/mask_heads/fcn_mask_head.py", line 111, in get_target gt_masks, rcnn_train_cfg) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 12, in mask_target mask_targets = torch.cat(list(mask_targets)) File "/home/pc/dev/mmdetection/mmdet/core/mask/mask_target.py", line 22, in mask_target_single _, maxh, maxw = gt_masks.shape AttributeError: 'list' object has no attribute 'shape'

10,603

open-mmlab/mmdetection

open-mmlab__mmdetection-2492

c1ef12df9c9d35f1402734435b23a4ae711f3084

diff --git a/mmdet/datasets/pipelines/transforms.py b/mmdet/datasets/pipelines/transforms.py --- a/mmdet/datasets/pipelines/transforms.py +++ b/mmdet/datasets/pipelines/transforms.py @@ -490,6 +490,9 @@ def __init__(self, def __call__(self, results): img = results['img'] + assert img.dtype == np.float32, \ + 'PhotoMetricDistortion needs the input image of dtype np.float32,'\ + ' please set "to_float32=True" in "LoadImageFromFile" pipeline' # random brightness if random.randint(2): delta = random.uniform(-self.brightness_delta,

_pickle.PicklingError: Can't pickle <class 'numpy.core._exceptions.UFuncTypeError'>: it's not the same object as numpy.core._exceptions.UFuncTypeError when i use dict(type='PhotoMetricDistortion', brightness_delta=32, contrast_range=(0.5, 1.5), saturation_range=(0.5, 1.5), hue_delta=18) in config file for training i got the following error: Traceback (most recent call last): File "/usr/lib/python3.6/multiprocessing/queues.py", line 234, in _feed obj = _ForkingPickler.dumps(obj) File "/usr/lib/python3.6/multiprocessing/reduction.py", line 51, in dumps cls(buf, protocol).dump(obj) _pickle.PicklingError: Can't pickle <class 'numpy.core._exceptions.UFuncTypeError'>: it's not the same object as numpy.core._exceptions.UFuncTypeError that's all info. i found the following code raise the error: delta = random.uniform(-self.brightness_delta,self.brightness_delta) img+=delta in PhotoMetricDistortion class in pipelines/transforms.py file other familiar code raise the error too. i don't know why? i found if delta is negative or float,it will raise the error.positive integer is ok.

when i use img = img + delta to replace img+=delta,it's ok... Do you use the config [ssd300_coco.py](https://github.com/open-mmlab/mmdetection/blob/master/configs/ssd300_coco.py), I can't reproduce your error with this config. Could you specify which config you use and what modification you apply to? Could you provide more detail information about your environment? when i use PhotoMetricDistortion in retinanet training, I met exactly same error > Do you use the config ssd300_coco.py, I can't reproduce your error with this config. Could you specify which config you use and what modification you apply to? Could you provide more detail information about your environment? I use the faster_rcnn_r50_fpn. torch-1.1.0, CUDA-10.1, numpy-1.18.2.Thanks. As mentioned above, when i replace '+=','*=',and such operation in PhotoMetricDistortion, the error disappear. I found that the problem is that I didn't load image with float32=True, cv2.cvtColor doesn't support float64. That causes this wired error here:) > > Do you use the config ssd300_coco.py, I can't reproduce your error with this config. Could you specify which config you use and what modification you apply to? Could you provide more detail information about your environment? > > I use the faster_rcnn_r50_fpn. torch-1.1.0, CUDA-10.1, numpy-1.18.2.Thanks. As mentioned above, when i replace '+=','*=',and such operation in PhotoMetricDistortion, the error disappear. @BChunlei Just as @edwardyangxin mentioned, when using PhotoMetricDistortion transformation, it is necessary to convert the image to np.float32 firstly just as what SSD does, you can refer [here](https://github.com/open-mmlab/mmdetection/blob/365c9302ee1eb5790b8e57f24ea4dfee8f2b88ac/configs/ssd300_coco.py#L51). But it is indeed not very friendly, would you like to create a PR to make it more friendly, i.e., report a reminder that user needs to set `to_float32` flag?

2020-04-20T14:48:29Z

[]

Traceback (most recent call last): File "/usr/lib/python3.6/multiprocessing/queues.py", line 234, in _feed obj = _ForkingPickler.dumps(obj) File "/usr/lib/python3.6/multiprocessing/reduction.py", line 51, in dumps cls(buf, protocol).dump(obj) _pickle.PicklingError: Can't pickle <class 'numpy.core._exceptions.UFuncTypeError'>: it's not the same object as numpy.core._exceptions.UFuncTypeError

10,612

open-mmlab/mmdetection

open-mmlab__mmdetection-2824

24a43d5060adb1b523a018eeee17d5ca75b2b23a

diff --git a/mmdet/core/mask/__init__.py b/mmdet/core/mask/__init__.py --- a/mmdet/core/mask/__init__.py +++ b/mmdet/core/mask/__init__.py @@ -1,7 +1,8 @@ from .mask_target import mask_target from .structures import BitmapMasks, PolygonMasks -from .utils import split_combined_polys +from .utils import encode_mask_results, split_combined_polys __all__ = [ - 'split_combined_polys', 'mask_target', 'BitmapMasks', 'PolygonMasks' + 'split_combined_polys', 'mask_target', 'BitmapMasks', 'PolygonMasks', + 'encode_mask_results' ] diff --git a/mmdet/core/mask/utils.py b/mmdet/core/mask/utils.py --- a/mmdet/core/mask/utils.py +++ b/mmdet/core/mask/utils.py @@ -1,4 +1,6 @@ import mmcv +import numpy as np +import pycocotools.mask as mask_util def split_combined_polys(polys, poly_lens, polys_per_mask): @@ -28,3 +30,34 @@ def split_combined_polys(polys, poly_lens, polys_per_mask): mask_polys = mmcv.slice_list(split_polys, polys_per_mask_single) mask_polys_list.append(mask_polys) return mask_polys_list + + +# TODO: move this function to more proper place +def encode_mask_results(mask_results): + """Encode bitmap mask to RLE code. + + Args: + mask_results (list | tuple[list]): bitmap mask results. + In mask scoring rcnn, mask_results is a tuple of (segm_results, + segm_cls_score). + + Returns: + list | tuple: RLE encoded mask. + """ + if isinstance(mask_results, tuple): # mask scoring + cls_segms, cls_mask_scores = mask_results + else: + cls_segms = mask_results + num_classes = len(cls_segms) + encoded_mask_results = [[] for _ in range(num_classes)] + for i in range(len(cls_segms)): + for cls_segm in cls_segms[i]: + encoded_mask_results[i].append( + mask_util.encode( + np.array( + cls_segm[:, :, np.newaxis], order='F', + dtype='uint8'))[0]) # encoded with RLE + if isinstance(mask_results, tuple): + return encoded_mask_results, cls_mask_scores + else: + return encoded_mask_results diff --git a/mmdet/models/detectors/base.py b/mmdet/models/detectors/base.py --- a/mmdet/models/detectors/base.py +++ b/mmdet/models/detectors/base.py @@ -3,7 +3,6 @@ import mmcv import numpy as np -import pycocotools.mask as maskUtils import torch.nn as nn from mmcv.utils import print_log @@ -210,7 +209,7 @@ def show_result(self, for i in inds: i = int(i) color_mask = color_masks[labels[i]] - mask = maskUtils.decode(segms[i]).astype(np.bool) + mask = segms[i] img[mask] = img[mask] * 0.5 + color_mask * 0.5 # if out_file specified, do not show image in window if out_file is not None: diff --git a/mmdet/models/detectors/cascade_rcnn.py b/mmdet/models/detectors/cascade_rcnn.py --- a/mmdet/models/detectors/cascade_rcnn.py +++ b/mmdet/models/detectors/cascade_rcnn.py @@ -31,4 +31,4 @@ def show_result(self, data, result, **kwargs): else: if isinstance(result, dict): result = result['ensemble'] - super(CascadeRCNN, self).show_result(data, result, **kwargs) + return super(CascadeRCNN, self).show_result(data, result, **kwargs)

IndexError in pycocotools Thanks for your error report and we appreciate it a lot. **Describe the bug** When running the image_demo.py, I get an error about numpy index. **Reproduction** ``` python demo/image_demo.py demo/demo.jpg configs/mask_rcnn/mask_rcnn_r50_fpn_1x_coco.py checkpoints/mask_rcnn_r50_fpn_1x_coco_20200205-d4b0c5d6.pth --device cpu ``` **Environment** `ys.platform: linux Python: 3.7.4 (default, Aug 13 2019, 20:35:49) [GCC 7.3.0] CUDA available: True CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 10.0, V10.0.130 GPU 0: GeForce GTX 1080 Ti GCC: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609 PyTorch: 1.4.0 PyTorch compiling details: PyTorch built with: - GCC 7.3 - Intel(R) Math Kernel Library Version 2019.0.4 Product Build 20190411 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v0.21.1 (Git Hash 7d2fd500bc78936d1d648ca713b901012f470dbc) - OpenMP 201511 (a.k.a. OpenMP 4.5) - NNPACK is enabled - CUDA Runtime 10.1 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37 - CuDNN 7.6.3 - Magma 2.5.1 - Build settings: BLAS=MKL, BUILD_NAMEDTENSOR=OFF, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -O2 -fPIC -Wno-narrowing -Wall -Wextra -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Wno-stringop-overflow, DISABLE_NUMA=1, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_STATIC_DISPATCH=OFF, TorchVision: 0.5.0 OpenCV: 4.2.0 MMCV: 0.5.4 MMDetection: 2.0.0+unknown MMDetection Compiler: GCC 5.4 MMDetection CUDA Compiler: 10.0 **Error traceback** ``` Traceback (most recent call last): File "demo/image_demo.py", line 26, in <module> main() File "demo/image_demo.py", line 22, in main show_result_pyplot(model, args.img, result, score_thr=args.score_thr) File "/root/mmdetection-master/mmdet/apis/inference.py", line 146, in show_result_pyplot img = model.show_result(img, result, score_thr=score_thr, show=False) File "/root/mmdetection-master/mmdet/models/detectors/base.py", line 211, in show_result mask = maskUtils.decode(segms[i]).astype(np.bool) File "/root/anaconda3/lib/python3.7/site-packages/pycocotools-2.0-py3.7-linux-x86_64.egg/pycocotools/mask.py", line 91, in decode return _mask.decode([rleObjs])[:,:,0] File "pycocotools/_mask.pyx", line 146, in pycocotools._mask.decode File "pycocotools/_mask.pyx", line 128, in pycocotools._mask._frString IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices ```

Please try to install pycocotools through pip: `pip install "git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI"`. Sometimes the numpy version can also cause some problem. If the error still exist when you have installed the right pycocotools, you may try numpy=1.17 @Johnson-Wang I did install pycocotools from github, and the numpy version is 1.17, but the problem is not solved. @Johnson-Wang and I also tried lowering the numpy version, it just doesn't work Maybe try numpy>=1.18 ? @ZwwWayne Yes, I also try numpy 1.18, but it still doesn't work Did you install pycocotools from pip before you install from github using `pip install "git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI"`, it might because the environment already have one and the new installation does not work. In this case you need to uninstall all pycocotools in your environment and re-install a new one. @ZwwWayne Before install from github, I uninstalled pycocotools and make sure that there is no pycocotools libarary files in the system, but it just can't work. Then how about numpy? Did you do the same thing? Yes, I did everythingk, I even uninstall conda and clean all python environment, it doesn't work In object detection, try to rewrite the `def show_result(self, data, result, **kwargs) `in your detector.py and add `self.CLASSES = tuple('A','B', ... , 'Z')` before `super(detector_name, self).show_result(data, result, **kwargs)`. There may be similar labels in segm @aimhabo Actually I am doing something about instance segmentation, so I have to call the pycocotools , otherwise I will delete the segmentation part. But I don't understand how could similar labels cause an error in numpy? @mangdian I mean the similar ground-truth information like labels. In the dataset settings, it will load MS-COCO's information automatically while you use other dataset with `dataset_type = 'CocoDataset'`. In my similar situation, I force the detector's `self.CLASSES` in `show_result()` (In __init__() the forcement is failed because of any rewriting I haven't found yet). The problem is that the segmentation returned by the models is numpy boolean arrays instead of undecoded bytes data, so there is no need to decode the result. I finally solve this problem by modifying mmdet/apis/inference.py (delete the decode part). However, it is very strange that the code works well on another linux server of mine, no need to modify inference.py. Anyway, modifying inference.py can be a temporal solution to solve this problem. Hi @mangdian , Thanks for your bug report. Do you use the newest version when you meet the bug? It seems that we did not change inference.py after changing the test logic. Would you like to create a PR to fix that? @ZwwWayne Yes, I am using the newest version of mmdetection. What confuses me is that I tried the demo in two linux server. On one machine the demo code works, while on the other it doesn't work. I am not sure if it is a bug, or it is something related the numpy version or Cython version (if it is related to version, modifying the code may not be a good choice). But the reason that leads to the error is clear. The segmentation result returned by the model is numpy boolean array, insead of encoded bytes, so the cocotools failed to decode the results. It seems that some other people also meet this problem. Should I create a PR? Met the same problem here. I install the newest version mmdetection following the guide, it very annoying to met this error while testing`mask rcnn` with the demo code: ``` IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices ``` What's more, the model download urls for `aliyun` was not work @FantasyJXF Please try my solution: replace line 213 in mmdet/models/detectors/base.py: ` mask = maskUtils.decode(segms[i]).astype(np.bool) ` with `mask = segms[i]` > @FantasyJXF Please try my solution: > > replace line 213 in mmdet/models/detectors/base.py: > ` mask = maskUtils.decode(segms[i]).astype(np.bool)` > with > `mask = segms[i]` Very useful, thanks for your contribution. So there is no need to do the extra decode to get boolean array, the model already do that, maybe that's because the developer didn't check the segmentation model while release. First day using mmdetection, almost turn to dedectron2. Thank you again. @mangdian Thanks for the fix, it will be appreciated if you could create a PR for it. @hellock OK， I will create a PR for it. Glad to do it. Two things to fix: 1. Change https://github.com/open-mmlab/mmdetection/blob/master/mmdet/models/detectors/base.py#L213 to `mask = segms[i]`. 2. Change https://github.com/open-mmlab/mmdetection/blob/master/mmdet/models/detectors/cascade_rcnn.py#L34 to `return xxxx` #2697 @hellock Hi, anthor not related problem in [model zoo](https://github.com/open-mmlab/mmdetection/blob/master/docs/model_zoo.md) It seems the mirror sites for ALIYUN was not work, and the AWS mirror url postfix `open-mmlab` seems changed to `mmdetection`. Yes, the aliyun mirror site has not been updated to host v2.0 models. It will be updated later.

2020-05-27T02:14:48Z

[]

Traceback (most recent call last): File "demo/image_demo.py", line 26, in <module> main() File "demo/image_demo.py", line 22, in main show_result_pyplot(model, args.img, result, score_thr=args.score_thr) File "/root/mmdetection-master/mmdet/apis/inference.py", line 146, in show_result_pyplot img = model.show_result(img, result, score_thr=score_thr, show=False) File "/root/mmdetection-master/mmdet/models/detectors/base.py", line 211, in show_result mask = maskUtils.decode(segms[i]).astype(np.bool) File "/root/anaconda3/lib/python3.7/site-packages/pycocotools-2.0-py3.7-linux-x86_64.egg/pycocotools/mask.py", line 91, in decode return _mask.decode([rleObjs])[:,:,0] File "pycocotools/_mask.pyx", line 146, in pycocotools._mask.decode File "pycocotools/_mask.pyx", line 128, in pycocotools._mask._frString IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices

10,617

open-mmlab/mmdetection

open-mmlab__mmdetection-3529

ae453fa92ffebcbd224b72f6d48e0b8699424450

diff --git a/tools/fuse_conv_bn.py b/tools/fuse_conv_bn.py deleted file mode 100644 --- a/tools/fuse_conv_bn.py +++ /dev/null @@ -1,67 +0,0 @@ -import argparse - -import torch -import torch.nn as nn -from mmcv.runner import save_checkpoint - -from mmdet.apis import init_detector - - -def fuse_conv_bn(conv, bn): - """During inference, the functionary of batch norm layers is turned off but - only the mean and var alone channels are used, which exposes the chance to - fuse it with the preceding conv layers to save computations and simplify - network structures.""" - conv_w = conv.weight - conv_b = conv.bias if conv.bias is not None else torch.zeros_like( - bn.running_mean) - - factor = bn.weight / torch.sqrt(bn.running_var + bn.eps) - conv.weight = nn.Parameter(conv_w * - factor.reshape([conv.out_channels, 1, 1, 1])) - conv.bias = nn.Parameter((conv_b - bn.running_mean) * factor + bn.bias) - return conv - - -def fuse_module(m): - last_conv = None - last_conv_name = None - - for name, child in m.named_children(): - if isinstance(child, (nn.BatchNorm2d, nn.SyncBatchNorm)): - if last_conv is None: # only fuse BN that is after Conv - continue - fused_conv = fuse_conv_bn(last_conv, child) - m._modules[last_conv_name] = fused_conv - # To reduce changes, set BN as Identity instead of deleting it. - m._modules[name] = nn.Identity() - last_conv = None - elif isinstance(child, nn.Conv2d): - last_conv = child - last_conv_name = name - else: - fuse_module(child) - return m - - -def parse_args(): - parser = argparse.ArgumentParser( - description='fuse Conv and BN layers in a model') - parser.add_argument('config', help='config file path') - parser.add_argument('checkpoint', help='checkpoint file path') - parser.add_argument('out', help='output path of the converted model') - args = parser.parse_args() - return args - - -def main(): - args = parse_args() - # build the model from a config file and a checkpoint file - model = init_detector(args.config, args.checkpoint) - # fuse conv and bn layers of the model - fused_model = fuse_module(model) - save_checkpoint(fused_model, args.out) - - -if __name__ == '__main__': - main()

ModuleNotFoundError: No module named 'tools' i would like to test the result of training, so i write the next: (base) zhangshen@zhangshen-X550JX:~/mmdetection$ python tools/test.py configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py checkpoints/faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth --out./result/result_100/pkl --eval bbox but i got: Traceback (most recent call last): File "tools/test.py", line 9, in <module> from tools.fuse_conv_bn import fuse_module ModuleNotFoundError: No module named 'tools' how can i solve this problem?

#2667. @yhcao6 Either fix the tool script or update the documentation. > @yhcao6 Either fix the tool script or update the documentation. how can i fix the tool script or which doucumentation should i update? I have created a pr to fix the error-report template, #3192 . Thanks for reporting the error.

2020-08-11T08:38:52Z

[]

Traceback (most recent call last): File "tools/test.py", line 9, in <module> from tools.fuse_conv_bn import fuse_module ModuleNotFoundError: No module named 'tools'

10,626

open-mmlab/mmdetection

open-mmlab__mmdetection-3836

8e29b954a0276593fae3469eaf0d205da145c3da

diff --git a/mmdet/models/dense_heads/reppoints_head.py b/mmdet/models/dense_heads/reppoints_head.py --- a/mmdet/models/dense_heads/reppoints_head.py +++ b/mmdet/models/dense_heads/reppoints_head.py @@ -292,7 +292,7 @@ def forward_single(self, x): pts_out_refine = pts_out_refine + pts_out_init.detach() return cls_out, pts_out_init, pts_out_refine - def get_points(self, featmap_sizes, img_metas): + def get_points(self, featmap_sizes, img_metas, device): """Get points according to feature map sizes. Args: @@ -310,7 +310,7 @@ def get_points(self, featmap_sizes, img_metas): multi_level_points = [] for i in range(num_levels): points = self.point_generators[i].grid_points( - featmap_sizes[i], self.point_strides[i]) + featmap_sizes[i], self.point_strides[i], device) multi_level_points.append(points) points_list = [[point.clone() for point in multi_level_points] for _ in range(num_imgs)] @@ -326,7 +326,7 @@ def get_points(self, featmap_sizes, img_metas): valid_feat_h = min(int(np.ceil(h / point_stride)), feat_h) valid_feat_w = min(int(np.ceil(w / point_stride)), feat_w) flags = self.point_generators[i].valid_flags( - (feat_h, feat_w), (valid_feat_h, valid_feat_w)) + (feat_h, feat_w), (valid_feat_h, valid_feat_w), device) multi_level_flags.append(flags) valid_flag_list.append(multi_level_flags) @@ -534,6 +534,7 @@ def loss_single(self, cls_score, pts_pred_init, pts_pred_refine, labels, label_weights = label_weights.reshape(-1) cls_score = cls_score.permute(0, 2, 3, 1).reshape(-1, self.cls_out_channels) + cls_score = cls_score.contiguous() loss_cls = self.loss_cls( cls_score, labels, @@ -572,11 +573,12 @@ def loss(self, gt_bboxes_ignore=None): featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores] assert len(featmap_sizes) == len(self.point_generators) + device = cls_scores[0].device label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1 # target for initial stage center_list, valid_flag_list = self.get_points(featmap_sizes, - img_metas) + img_metas, device) pts_coordinate_preds_init = self.offset_to_pts(center_list, pts_preds_init) if self.train_cfg.init.assigner['type'] == 'PointAssigner': @@ -604,7 +606,7 @@ def loss(self, # target for refinement stage center_list, valid_flag_list = self.get_points(featmap_sizes, - img_metas) + img_metas, device) pts_coordinate_preds_refine = self.offset_to_pts( center_list, pts_preds_refine) bbox_list = [] @@ -666,6 +668,7 @@ def get_bboxes(self, rescale=False, nms=True): assert len(cls_scores) == len(pts_preds_refine) + device = cls_scores[0].device bbox_preds_refine = [ self.points2bbox(pts_pred_refine) for pts_pred_refine in pts_preds_refine @@ -673,7 +676,7 @@ def get_bboxes(self, num_levels = len(cls_scores) mlvl_points = [ self.point_generators[i].grid_points(cls_scores[i].size()[-2:], - self.point_strides[i]) + self.point_strides[i], device) for i in range(num_levels) ] result_list = []

RuntimeError: expected device cuda:1 but got device cuda:0 I have 2 titan xp gpus ,and when I run reppoints detection train.py it occurs: ``` python tools/train.py workproject/gureppoints/reppoints_moment_r101_fpn_gn-neck+head_2x_coco.py --gpu-ids=1 2020-09-24 17:04:35,811 - mmdet - INFO - Distributed training: False 2020-09-24 17:04:36,261 - mmdet - INFO - Config: dataset_type = 'CocoDataset' data_root = 'data/coco/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(2048, 1024), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(2048, 1024), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=4, train=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_train2017.json', img_prefix='data/coco/train2017/', pipeline=[ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(2048, 1024), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ]), val=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_val2017.json', img_prefix='data/coco/val2017/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(2048, 1024), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ]), test=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_test2017.json', img_prefix='data/coco/test2017/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(2048, 1024), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ])) evaluation = dict(interval=1, metric='bbox') optimizer = dict(type='SGD', lr=0.00125, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=None) lr_config = dict( policy='step', warmup='linear', warmup_iters=500, warmup_ratio=0.001, step=[16, 22]) total_epochs = 24 checkpoint_config = dict(interval=1) log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')]) dist_params = dict(backend='nccl') log_level = 'INFO' load_from = None resume_from = None workflow = [('train', 1)] model = dict( type='RepPointsDetector', pretrained='torchvision://resnet101', backbone=dict( type='ResNet', depth=101, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, style='pytorch'), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, start_level=1, add_extra_convs='on_input', num_outs=5, norm_cfg=dict(type='GN', num_groups=32, requires_grad=True)), bbox_head=dict( type='RepPointsHead', num_classes=6, in_channels=256, feat_channels=256, point_feat_channels=256, stacked_convs=3, num_points=9, gradient_mul=0.1, point_strides=[8, 16, 32, 64, 128], point_base_scale=4, loss_cls=dict( type='FocalLoss', use_sigmoid=True, gamma=2.0, alpha=0.25, loss_weight=1.0), loss_bbox_init=dict(type='SmoothL1Loss', beta=0.11, loss_weight=0.5), loss_bbox_refine=dict(type='SmoothL1Loss', beta=0.11, loss_weight=1.0), transform_method='moment', norm_cfg=dict(type='GN', num_groups=32, requires_grad=True))) train_cfg = dict( init=dict( assigner=dict(type='PointAssigner', scale=4, pos_num=1), allowed_border=-1, pos_weight=-1, debug=False), refine=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.4, min_pos_iou=0, ignore_iof_thr=-1), allowed_border=-1, pos_weight=-1, debug=False)) test_cfg = dict( nms_pre=1000, min_bbox_size=0, score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100) norm_cfg = dict(type='GN', num_groups=32, requires_grad=True) work_dir = 'workproject/gureppoints' gpu_ids = [1] 2020-09-24 17:04:37,362 - mmdet - INFO - load model from: torchvision://resnet101 2020-09-24 17:04:38,153 - mmdet - WARNING - The model and loaded state dict do not match exactly unexpected key in source state_dict: fc.weight, fc.bias loading annotations into memory... Done (t=0.29s) creating index... index created! loading annotations into memory... Done (t=0.06s) creating index... index created! 2020-09-24 17:04:44,095 - mmdet - INFO - Start running, host: ys@ys, work_dir: /media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/workproject/gureppoints 2020-09-24 17:04:44,096 - mmdet - INFO - workflow: [('train', 1)], max: 24 epochs Traceback (most recent call last): File "tools/train.py", line 178, in <module> main() File "tools/train.py", line 174, in main meta=meta) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/apis/train.py", line 143, in train_detector runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 122, in run epoch_runner(data_loaders[i], **kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 32, in train **kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(*inputs[0], **kwargs[0]) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 234, in train_step losses = self(**data) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/core/fp16/decorators.py", line 51, in new_func return old_func(*args, **kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 168, in forward return self.forward_train(img, img_metas, **kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/single_stage.py", line 94, in forward_train gt_labels, gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 581, in loss pts_preds_init) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 368, in offset_to_pts pts = xy_pts_shift * self.point_strides[i_lvl] + pts_center RuntimeError: expected device cuda:1 but got device cuda:0 ``` so what's the problem?

Not typo. `print(args)` shows `gpu_ids=[1]`, and the log shows `gpu_ids = [1]`. > Not typo. > `print(args)` shows `gpu_ids=[1]`, and the log shows `gpu_ids = [1]`. you are right，I don't understand the usage of argparse well Does passing `device = 'cuda:1'` from `reppoints_head.py` work? `points = self.point_generators[i].grid_points(featmap_sizes[i], self.point_strides[i], device=device)` `flags = self.point_generators[i].valid_flags((feat_h, feat_w), (valid_feat_h, valid_feat_w), device=device)` If so, `pts = xy_pts_shift * self.point_strides[i_lvl] + pts_center.to(xy_pts_shift.device)` would be a tentative fix.

2020-09-25T07:15:13Z

[]

Traceback (most recent call last): File "tools/train.py", line 178, in <module> main() File "tools/train.py", line 174, in main meta=meta) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/apis/train.py", line 143, in train_detector runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 122, in run epoch_runner(data_loaders[i], **kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 32, in train **kwargs) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(*inputs[0], **kwargs[0]) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 234, in train_step losses = self(**data) File "/home/ys/anaconda3/envs/tensorflow1/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/core/fp16/decorators.py", line 51, in new_func return old_func(*args, **kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/base.py", line 168, in forward return self.forward_train(img, img_metas, **kwargs) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/detectors/single_stage.py", line 94, in forward_train gt_labels, gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 581, in loss pts_preds_init) File "/media/ys/4E2CD69B656E8D93/guchenliang/mmdetection/mmdet/models/dense_heads/reppoints_head.py", line 368, in offset_to_pts pts = xy_pts_shift * self.point_strides[i_lvl] + pts_center RuntimeError: expected device cuda:1 but got device cuda:0

10,635

open-mmlab/mmdetection

open-mmlab__mmdetection-4555

3620bb74893ef688b3243652042660e1b5866d5e

diff --git a/mmdet/datasets/xml_style.py b/mmdet/datasets/xml_style.py --- a/mmdet/datasets/xml_style.py +++ b/mmdet/datasets/xml_style.py @@ -20,6 +20,8 @@ class XMLDataset(CustomDataset): """ def __init__(self, min_size=None, **kwargs): + assert self.CLASSES or kwargs.get( + 'classes', None), 'CLASSES in `XMLDataset` can not be None.' super(XMLDataset, self).__init__(**kwargs) self.cat2label = {cat: i for i, cat in enumerate(self.CLASSES)} self.min_size = min_size @@ -43,8 +45,6 @@ def load_annotations(self, ann_file): tree = ET.parse(xml_path) root = tree.getroot() size = root.find('size') - width = 0 - height = 0 if size is not None: width = int(size.find('width').text) height = int(size.find('height').text)

TypeError: argument of type 'NoneType' is not iterable When I use my own datasets to train faster-rcnn,I meet this question,the following are environment infomation and logs: sys.platform: linux Python: 3.7.9 | packaged by conda-forge | (default, Dec 9 2020, 21:08:20) [GCC 9.3.0] CUDA available: True GPU 0,1,2,3,4,5,6,7: GeForce GTX 1080 Ti CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 10.2, V10.2.89 GCC: gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 PyTorch: 1.5.0 PyTorch compiling details: PyTorch built with: - GCC 7.3 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v0.21.1 (Git Hash 7d2fd500bc78936d1d648ca713b901012f470dbc) - OpenMP 201511 (a.k.a. OpenMP 4.5) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_6 1;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37 - CuDNN 7.6.5 - Magma 2.5.2 - Build settings: BLAS=MKL, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -DNDEBUG -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DUSE_INTERNAL_THREADPOOL_IMPL -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_STATIC_DISPATCH=OFF, TorchVision: 0.6.0a0+82fd1c8 OpenCV: 4.4.0 MMCV: 1.2.4 MMCV Compiler: GCC 7.3 MMCV CUDA Compiler: 10.2 MMDetection: 2.8.0+ ------------------------------------------------------------ 2021-01-31 22:13:23,946 - mmdet - INFO - Distributed training: False 2021-01-31 22:13:26,538 - mmdet - INFO - Config: model = dict( type='FasterRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, style='pytorch'), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, feat_channels=256, anchor_generator=dict( type='AnchorGenerator', scales=[8], ratios=[0.5, 1.0, 2.0], strides=[4, 8, 16, 32, 64]), bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[1.0, 1.0, 1.0, 1.0]), loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0)), roi_head=dict( type='StandardRoIHead', bbox_roi_extractor=dict( type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0), out_channels=256, featmap_strides=[4, 8, 16, 32]), bbox_head=dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=232, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.1, 0.1, 0.2, 0.2]), reg_class_agnostic=False, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0))), train_cfg=dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, match_low_quality=True, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=-1, pos_weight=-1, debug=False), rpn_proposal=dict( nms_across_levels=False, nms_pre=2000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False)), test_cfg=dict( rpn=dict( nms_across_levels=False, nms_pre=1000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100))) dataset_type = 'VOCDataset' data_root = '/home/chengyuhong/mmdetection/data/tt100k_2021/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=2, train=dict( type='RepeatDataset', times=3, dataset=dict( type='VOCDataset', ann_file=[ '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/ImageSets/Main/trainval.txt' ], img_prefix=[ '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/' ], pipeline=[ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ])), val=dict( type='VOCDataset', ann_file= '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/ImageSets/Main/test.txt', img_prefix='/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ]), test=dict( type='VOCDataset', ann_file= '/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/ImageSets/Main/test.txt', img_prefix='/home/chengyuhong/mmdetection/data/tt100k_2021/VOC2007/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ])) evaluation = dict(interval=1, metric='mAP') checkpoint_config = dict(interval=1) log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')]) custom_hooks = [dict(type='NumClassCheckHook')] dist_params = dict(backend='nccl') log_level = 'INFO' load_from = None resume_from = None workflow = [('train', 1)] optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=None) lr_config = dict(policy='step', step=[3]) total_epochs = 4 work_dir = './work_dirs/faster_rcnn_r50_fpn_1x_voc0712' gpu_ids = range(0, 1) 2021-01-31 22:13:27,410 - mmdet - INFO - load model from: torchvision://resnet50 2021-01-31 22:13:27,935 - mmdet - WARNING - The model and loaded state dict do not match exactly unexpected key in source state_dict: fc.weight, fc.bias Traceback (most recent call last): File "./tools/train.py", line 187, in <module> main() File "./tools/train.py", line 163, in main datasets = [build_dataset(cfg.data.train)] File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 64, in build_dataset build_dataset(cfg['dataset'], default_args), cfg['times']) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 69, in build_dataset dataset = _concat_dataset(cfg, default_args) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 48, in _concat_dataset datasets.append(build_dataset(data_cfg, default_args)) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 71, in build_dataset dataset = build_from_cfg(cfg, DATASETS, default_args) File "/home/chengyuhong/anaconda3/envs/open-mmlab/lib/python3.7/site-packages/mmcv/utils/registry.py", line 171, in build_from_cfg return obj_cls(**args) File "/home/chengyuhong/TT100K/mmdet/datasets/voc.py", line 32, in __init__ super(VOCDataset, self).__init__(**kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 23, in __init__ super(XMLDataset, self).__init__(**kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/custom.py", line 96, in __init__ valid_inds = self._filter_imgs() File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 75, in _filter_imgs if name in self.CLASSES: TypeError: argument of type 'NoneType' is not iterable

2021-01-31T14:25:14Z

[]

Traceback (most recent call last): File "./tools/train.py", line 187, in <module> main() File "./tools/train.py", line 163, in main datasets = [build_dataset(cfg.data.train)] File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 64, in build_dataset build_dataset(cfg['dataset'], default_args), cfg['times']) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 69, in build_dataset dataset = _concat_dataset(cfg, default_args) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 48, in _concat_dataset datasets.append(build_dataset(data_cfg, default_args)) File "/home/chengyuhong/TT100K/mmdet/datasets/builder.py", line 71, in build_dataset dataset = build_from_cfg(cfg, DATASETS, default_args) File "/home/chengyuhong/anaconda3/envs/open-mmlab/lib/python3.7/site-packages/mmcv/utils/registry.py", line 171, in build_from_cfg return obj_cls(**args) File "/home/chengyuhong/TT100K/mmdet/datasets/voc.py", line 32, in __init__ super(VOCDataset, self).__init__(**kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 23, in __init__ super(XMLDataset, self).__init__(**kwargs) File "/home/chengyuhong/TT100K/mmdet/datasets/custom.py", line 96, in __init__ valid_inds = self._filter_imgs() File "/home/chengyuhong/TT100K/mmdet/datasets/xml_style.py", line 75, in _filter_imgs if name in self.CLASSES: TypeError: argument of type 'NoneType' is not iterable

10,650

open-mmlab/mmdetection

open-mmlab__mmdetection-5654

475c6bec197e6b495d48146636301590165d8f66

diff --git a/mmdet/models/detectors/detr.py b/mmdet/models/detectors/detr.py --- a/mmdet/models/detectors/detr.py +++ b/mmdet/models/detectors/detr.py @@ -1,3 +1,5 @@ +import warnings + import torch from ..builder import DETECTORS @@ -19,6 +21,27 @@ def __init__(self, super(DETR, self).__init__(backbone, None, bbox_head, train_cfg, test_cfg, pretrained, init_cfg) + # over-write `forward_dummy` because: + # the forward of bbox_head requires img_metas + def forward_dummy(self, img): + """Used for computing network flops. + + See `mmdetection/tools/analysis_tools/get_flops.py` + """ + warnings.warn('Warning! MultiheadAttention in DETR does not ' + 'support flops computation! Do not use the ' + 'results in your papers!') + + batch_size, _, height, width = img.shape + dummy_img_metas = [ + dict( + batch_input_shape=(height, width), + img_shape=(height, width, 3)) for _ in range(batch_size) + ] + x = self.extract_feat(img) + outs = self.bbox_head(x, dummy_img_metas) + return outs + # over-write `onnx_export` because: # (1) the forward of bbox_head requires img_metas # (2) the different behavior (e.g. construction of `masks`) between

Error get params DETR/ Deformable DETR Despite my attempts to modify, also just testing with the basic config detr file. Maybe this issue has already been raised? mmdet==2.13.0 mmcv=1.3.3 ```python python tools/analysis_tools/get_flops.py configs/detr/detr_r50_8x2_150e_coco.py ``` ```python /home/bluav/mmdetection/mmdet/models/backbones/resnet.py:400: UserWarning: DeprecationWarning: pretrained is a deprecated, please use "init_cfg" instead warnings.warn('DeprecationWarning: pretrained is a deprecated, ' Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Warning: variables __flops__ or __params__ are already defined for the moduleReLU ptflops can affect your code! Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) File "/home/bluav/mmdetection/mmdet/models/detectors/single_stage.py", line 48, in forward_dummy outs = self.bbox_head(x) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) TypeError: forward() missing 1 required positional argument: 'img_metas' ```

2021-07-20T05:56:04Z

[]

Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) File "/home/bluav/mmdetection/mmdet/models/detectors/single_stage.py", line 48, in forward_dummy outs = self.bbox_head(x) File "/home/bluav/.conda/envs/open-mmlab/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) TypeError: forward() missing 1 required positional argument: 'img_metas'

10,660

open-mmlab/mmdetection

open-mmlab__mmdetection-5884

6882fa0a220880e6e2e7e0536037f043b9031185

diff --git a/mmdet/models/roi_heads/bbox_heads/bbox_head.py b/mmdet/models/roi_heads/bbox_heads/bbox_head.py --- a/mmdet/models/roi_heads/bbox_heads/bbox_head.py +++ b/mmdet/models/roi_heads/bbox_heads/bbox_head.py @@ -455,9 +455,15 @@ def regress_by_class(self, rois, label, bbox_pred, img_meta): """Regress the bbox for the predicted class. Used in Cascade R-CNN. Args: - rois (Tensor): shape (n, 4) or (n, 5) - label (Tensor): shape (n, ) - bbox_pred (Tensor): shape (n, 4*(#class)) or (n, 4) + rois (Tensor): Rois from `rpn_head` or last stage + `bbox_head`, has shape (num_proposals, 4) or + (num_proposals, 5). + label (Tensor): Only used when `self.reg_class_agnostic` + is False, has shape (num_proposals, ). + bbox_pred (Tensor): Regression prediction of + current stage `bbox_head`. When `self.reg_class_agnostic` + is False, it has shape (n, num_classes * 4), otherwise + it has shape (n, 4). img_meta (dict): Image meta info. Returns: diff --git a/mmdet/models/roi_heads/cascade_roi_head.py b/mmdet/models/roi_heads/cascade_roi_head.py --- a/mmdet/models/roi_heads/cascade_roi_head.py +++ b/mmdet/models/roi_heads/cascade_roi_head.py @@ -280,7 +280,28 @@ def forward_train(self, return losses def simple_test(self, x, proposal_list, img_metas, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ assert self.with_bbox, 'Bbox head must be implemented.' num_imgs = len(proposal_list) img_shapes = tuple(meta['img_shape'] for meta in img_metas) @@ -340,7 +361,7 @@ def simple_test(self, x, proposal_list, img_metas, rescale=False): if rois[j].shape[0] > 0: bbox_label = cls_score[j][:, :-1].argmax(dim=1) refined_rois = self.bbox_head[i].regress_by_class( - rois[j], bbox_label[j], bbox_pred[j], img_metas[j]) + rois[j], bbox_label, bbox_pred[j], img_metas[j]) refine_rois_list.append(refined_rois) rois = torch.cat(refine_rois_list) diff --git a/mmdet/models/roi_heads/htc_roi_head.py b/mmdet/models/roi_heads/htc_roi_head.py --- a/mmdet/models/roi_heads/htc_roi_head.py +++ b/mmdet/models/roi_heads/htc_roi_head.py @@ -326,7 +326,28 @@ def forward_train(self, return losses def simple_test(self, x, proposal_list, img_metas, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ if self.with_semantic: _, semantic_feat = self.semantic_head(x) else: @@ -381,7 +402,7 @@ def simple_test(self, x, proposal_list, img_metas, rescale=False): if rois[j].shape[0] > 0: bbox_label = cls_score[j][:, :-1].argmax(dim=1) refine_rois = bbox_head.regress_by_class( - rois[j], bbox_label[j], bbox_pred[j], img_metas[j]) + rois[j], bbox_label, bbox_pred[j], img_metas[j]) refine_rois_list.append(refine_rois) rois = torch.cat(refine_rois_list) diff --git a/mmdet/models/roi_heads/scnet_roi_head.py b/mmdet/models/roi_heads/scnet_roi_head.py --- a/mmdet/models/roi_heads/scnet_roi_head.py +++ b/mmdet/models/roi_heads/scnet_roi_head.py @@ -213,26 +213,19 @@ def forward_train(self, """ Args: x (list[Tensor]): list of multi-level img features. - img_metas (list[dict]): list of image info dict where each dict has: 'img_shape', 'scale_factor', 'flip', and may also contain 'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'. For details on the values of these keys see `mmdet/datasets/pipelines/formatting.py:Collect`. - proposal_list (list[Tensors]): list of region proposals. - gt_bboxes (list[Tensor]): Ground truth bboxes for each image with shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format. - gt_labels (list[Tensor]): class indices corresponding to each box - gt_bboxes_ignore (None, list[Tensor]): specify which bounding boxes can be ignored when computing the loss. - gt_masks (None, Tensor) : true segmentation masks for each box used if the architecture supports a segmentation task. - gt_semantic_seg (None, list[Tensor]): semantic segmentation masks used if the architecture supports semantic segmentation task. @@ -317,7 +310,28 @@ def forward_train(self, return losses def simple_test(self, x, proposal_list, img_metas, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ if self.with_semantic: _, semantic_feat = self.semantic_head(x) else: @@ -379,7 +393,7 @@ def simple_test(self, x, proposal_list, img_metas, rescale=False): if rois[j].shape[0] > 0: bbox_label = cls_score[j][:, :-1].argmax(dim=1) refine_rois = bbox_head.regress_by_class( - rois[j], bbox_label[j], bbox_pred[j], img_metas[j]) + rois[j], bbox_label, bbox_pred[j], img_metas[j]) refine_rois_list.append(refine_rois) rois = torch.cat(refine_rois_list) diff --git a/mmdet/models/roi_heads/standard_roi_head.py b/mmdet/models/roi_heads/standard_roi_head.py --- a/mmdet/models/roi_heads/standard_roi_head.py +++ b/mmdet/models/roi_heads/standard_roi_head.py @@ -224,7 +224,28 @@ def simple_test(self, img_metas, proposals=None, rescale=False): - """Test without augmentation.""" + """Test without augmentation. + + Args: + x (tuple[Tensor]): Features from upstream network. Each + has shape (batch_size, c, h, w). + proposal_list (list(Tensor)): Proposals from rpn head. + Each has shape (num_proposals, 5), last dimension + 5 represent (x1, y1, x2, y2, score). + img_metas (list[dict]): Meta information of images. + rescale (bool): Whether to rescale the results to + the original image. Default: True. + + Returns: + list[list[np.ndarray]] or list[tuple]: When no mask branch, + it is bbox results of each image and classes with type + `list[list[np.ndarray]]`. The outer list + corresponds to each image. The inner list + corresponds to each class. When the model has mask branch, + it contains bbox results and mask results. + The outer list corresponds to each image, and first element + of tuple is bbox results, second element is mask results. + """ assert self.with_bbox, 'Bbox head must be implemented.' det_bboxes, det_labels = self.simple_test_bboxes(

bbox_label dimension incorrect in cascade_roi_head.py simeple test **Describe the bug** I am trying the cascade rcnn network on a test dataset with a single image, after the training begins, the `regress_by_class` method called in `simple_test` of `cascade_roi_head.py` produce an error. **Error traceback** ``` 2021-08-07 12:48:22,062 - mmdet - INFO - Saving checkpoint at 1 epochs [ ] 0/1, elapsed: 0s Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 177, in main train_detector( File "c:\users\colli\mmdetection\mmdet\apis\train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 54, in train self.call_hook('after_train_epoch') File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\base_runner.py", line 307, in call_hook getattr(hook, fn_name)(self) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\hooks\evaluation.py", line 220, in after_train_epoch self._do_evaluate(runner) File "c:\users\colli\mmdetection\mmdet\core\evaluation\eval_hooks.py", line 17, in _do_evaluate results = single_gpu_test(runner.model, self.dataloader, show=False) File "c:\users\colli\mmdetection\mmdet\apis\test.py", line 27, in single_gpu_test result = model(return_loss=False, rescale=True, **data) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\parallel\data_parallel.py", line 42, in forward return super().forward(*inputs, **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\parallel\data_parallel.py", line 166, in forward return self.module(*inputs[0], **kwargs[0]) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 98, in new_func return old_func(*args, **kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 173, in forward return self.forward_test(img, img_metas, **kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 146, in forward_test return self.simple_test(imgs[0], img_metas[0], **kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\two_stage.py", line 181, in simple_test return self.roi_head.simple_test( File "c:\users\colli\mmdetection\mmdet\models\roi_heads\cascade_roi_head.py", line 344, in simple_test refined_rois = self.bbox_head[i].regress_by_class( File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 186, in new_func return old_func(*args, **kwargs) File "c:\users\colli\mmdetection\mmdet\models\roi_heads\bbox_heads\bbox_head.py", line 471, in regress_by_class inds = torch.stack((label, label + 1, label + 2, label + 3), 1) IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1) ``` **Environment** ``` sys.platform: win32 Python: 3.8.8 (default, Apr 13 2021, 15:08:03) [MSC v.1916 64 bit (AMD64)] CUDA available: True GPU 0: NVIDIA GeForce RTX 2060 SUPER CUDA_HOME: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.2 NVCC: Not Available GCC: n/a PyTorch: 1.9.0 PyTorch compiling details: PyTorch built with: - C++ Version: 199711 - MSVC 192829337 - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.1.2 (Git Hash 98be7e8afa711dc9b66c8ff3504129cb82013cdb) - OpenMP 2019 - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37 - CuDNN 7.6.5 - Magma 2.5.4 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=10.2, CUDNN_VERSION=7.6.5, CXX_COMPILER=C:/cb/pytorch_1000000000000/work/tmp_bin/sccache-cl.exe, CXX_FLAGS=/DWIN32 /D_WINDOWS /GR /EHsc /w /bigobj -DUSE_PTHREADPOOL -openmp:experimental -IC:/cb/pytorch_1000000000000/work/mkl/include -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.9.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=OFF, USE_NNPACK=OFF, USE_OPENMP=ON, TorchVision: 0.10.0 OpenCV: 4.5.2 MMCV: 1.3.9 MMCV Compiler: MSVC 192729111 MMCV CUDA Compiler: 11.2 MMDetection: 2.15.0+62a1cd3 ``` **Bug fix** https://github.com/open-mmlab/mmdetection/blob/62a1cd3fd6091fc4efa83627631ab4b99a8d005c/mmdet/models/roi_heads/cascade_roi_head.py#L332-L345 https://github.com/open-mmlab/mmdetection/blob/46988b3ac9820bcd8728980f04be98272ee5ea39/mmdet/models/roi_heads/bbox_heads/bbox_head.py#L453-L465 `regress_by_class` expects a shape (n, ), but in the code above, in line 341 bbox_label is produced by slicing the cls_score, so it has been a 1 dimension tensor, but in line 343, it is sliced again, so it seems to be a zero dimension scalar now. I guess this produce the problem, after changing it to ``` refined_rois = self.bbox_head[i].regress_by_class(rois[j], bbox_label, bbox_pred[j], img_metas[j]) ``` The code runs correctly. The change in these lines is made in fd5d019ca983b032c35d15cd900b0fa6eec4f988 by @hhaAndroid, can you have a look at it? Am I understanding it correctly?

@hhaAndroid can you pls take a look at it? Can anyone have a look at this issue?

2021-08-14T03:18:04Z

[]

Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 177, in main train_detector( File "c:\users\colli\mmdetection\mmdet\apis\train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 54, in train self.call_hook('after_train_epoch') File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\base_runner.py", line 307, in call_hook getattr(hook, fn_name)(self) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\hooks\evaluation.py", line 220, in after_train_epoch self._do_evaluate(runner) File "c:\users\colli\mmdetection\mmdet\core\evaluation\eval_hooks.py", line 17, in _do_evaluate results = single_gpu_test(runner.model, self.dataloader, show=False) File "c:\users\colli\mmdetection\mmdet\apis\test.py", line 27, in single_gpu_test result = model(return_loss=False, rescale=True, **data) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\parallel\data_parallel.py", line 42, in forward return super().forward(*inputs, **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\parallel\data_parallel.py", line 166, in forward return self.module(*inputs[0], **kwargs[0]) File "C:\Users\colli\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 98, in new_func return old_func(*args, **kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 173, in forward return self.forward_test(img, img_metas, **kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\base.py", line 146, in forward_test return self.simple_test(imgs[0], img_metas[0], **kwargs) File "c:\users\colli\mmdetection\mmdet\models\detectors\two_stage.py", line 181, in simple_test return self.roi_head.simple_test( File "c:\users\colli\mmdetection\mmdet\models\roi_heads\cascade_roi_head.py", line 344, in simple_test refined_rois = self.bbox_head[i].regress_by_class( File "C:\Users\colli\anaconda3\lib\site-packages\mmcv\runner\fp16_utils.py", line 186, in new_func return old_func(*args, **kwargs) File "c:\users\colli\mmdetection\mmdet\models\roi_heads\bbox_heads\bbox_head.py", line 471, in regress_by_class inds = torch.stack((label, label + 1, label + 2, label + 3), 1) IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1)

10,662

open-mmlab/mmdetection

open-mmlab__mmdetection-5930

0d2b39b195031b3b7ede73ebaa0a04150d1f332b

diff --git a/mmdet/datasets/pipelines/transforms.py b/mmdet/datasets/pipelines/transforms.py --- a/mmdet/datasets/pipelines/transforms.py +++ b/mmdet/datasets/pipelines/transforms.py @@ -2456,7 +2456,7 @@ def __call__(self, results): width = img.shape[1] + self.border[1] * 2 # Center - center_matrix = np.eye(3) + center_matrix = np.eye(3, dtype=np.float32) center_matrix[0, 2] = -img.shape[1] / 2 # x translation (pixels) center_matrix[1, 2] = -img.shape[0] / 2 # y translation (pixels) @@ -2561,21 +2561,24 @@ def __repr__(self): @staticmethod def _get_rotation_matrix(rotate_degrees): radian = math.radians(rotate_degrees) - rotation_matrix = np.array([[np.cos(radian), -np.sin(radian), 0.], - [np.sin(radian), - np.cos(radian), 0.], [0., 0., 1.]]) + rotation_matrix = np.array( + [[np.cos(radian), -np.sin(radian), 0.], + [np.sin(radian), np.cos(radian), 0.], [0., 0., 1.]], + dtype=np.float32) return rotation_matrix @staticmethod def _get_scaling_matrix(scale_ratio): - scaling_matrix = np.array([[scale_ratio, 0., 0.], - [0., scale_ratio, 0.], [0., 0., 1.]]) + scaling_matrix = np.array( + [[scale_ratio, 0., 0.], [0., scale_ratio, 0.], [0., 0., 1.]], + dtype=np.float32) return scaling_matrix @staticmethod def _get_share_matrix(scale_ratio): - scaling_matrix = np.array([[scale_ratio, 0., 0.], - [0., scale_ratio, 0.], [0., 0., 1.]]) + scaling_matrix = np.array( + [[scale_ratio, 0., 0.], [0., scale_ratio, 0.], [0., 0., 1.]], + dtype=np.float32) return scaling_matrix @staticmethod @@ -2583,10 +2586,12 @@ def _get_shear_matrix(x_shear_degrees, y_shear_degrees): x_radian = math.radians(x_shear_degrees) y_radian = math.radians(y_shear_degrees) shear_matrix = np.array([[1, np.tan(x_radian), 0.], - [np.tan(y_radian), 1, 0.], [0., 0., 1.]]) + [np.tan(y_radian), 1, 0.], [0., 0., 1.]], + dtype=np.float32) return shear_matrix @staticmethod def _get_translation_matrix(x, y): - translation_matrix = np.array([[1, 0., x], [0., 1, y], [0., 0., 1.]]) + translation_matrix = np.array([[1, 0., x], [0., 1, y], [0., 0., 1.]], + dtype=np.float32) return translation_matrix

RuntimeError: Expected object of scalar type Double but got scalar type Float for argument #2 'other' in call to _th_max when using mosaic mixup in cascade RCNN As mmdetection recently release 2.15.1, I want to use Mosaic and Mixup in Cascade RCNN, However, when I replace the relevant config from YOLOX to Cascade RCNN, I got the following error: ------------------------------------------------------------------------------------- 2021-08-12 06:53:24,512 - mmdet - INFO - workflow: [('train', 1)], max: 40 epochs Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 184, in main meta=meta) File "mmdetection-2.15.1/mmdet/apis/train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], **kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 50, in train self.run_iter(data_batch, train_mode=True, **kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 30, in run_iter **kwargs) File "mmcv-1.3.9/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(*inputs[0], **kwargs[0]) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 237, in train_step losses = self(**data) File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 98, in new_func return old_func(*args, **kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 171, in forward return self.forward_train(img, img_metas, **kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/two_stage.py", line 140, in forward_train proposal_cfg=proposal_cfg) File "mmdetection-2.15.1/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "mmdetection-2.15.1/mmdet/models/dense_heads/rpn_head.py", line 74, in loss gt_bboxes_ignore=gt_bboxes_ignore) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 186, in new_func return old_func(*args, **kwargs) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 463, in loss label_channels=label_channels) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 345, in get_targets unmap_outputs=unmap_outputs) File "mmdetection-2.15.1/mmdet/core/utils/misc.py", line 29, in multi_apply return tuple(map(list, zip(*map_results))) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 219, in _get_targets_single None if self.sampling else gt_labels) File "mmdetection-2.15.1/mmdet/core/bbox/assigners/max_iou_assigner.py", line 105, in assign overlaps = self.iou_calculator(gt_bboxes, bboxes) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 65, in __call__ return bbox_overlaps(bboxes1, bboxes2, mode, is_aligned) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 233, in bbox_overlaps bboxes2[..., None, :, :2]) # [B, rows, cols, 2] RuntimeError: Expected object of scalar type Double but got scalar type Float for argument #2 'other' in call to _th_max ------------------------------------------------------------------------------------- And below is my config: ------------------------------------------------------------------------------------- model = dict( type='CascadeRCNN', backbone=dict( type='ResNeXt', depth=101, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, style='pytorch', init_cfg=dict( type='Pretrained', checkpoint= 'pretrained_model/resnext101_64x4d-ee2c6f71.pth' ), groups=64, base_width=4), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, feat_channels=256, anchor_generator=dict( type='AnchorGenerator', scales=[8], ratios=[0.5, 1.0, 2.0], strides=[4, 8, 16, 32, 64]), bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[1.0, 1.0, 1.0, 1.0]), loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), loss_bbox=dict( type='SmoothL1Loss', beta=0.1111111111111111, loss_weight=1.0)), roi_head=dict( type='CascadeRoIHead', num_stages=3, stage_loss_weights=[1, 0.5, 0.25], bbox_roi_extractor=dict( type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0), out_channels=256, featmap_strides=[4, 8, 16, 32]), bbox_head=[ dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=12, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.1, 0.1, 0.2, 0.2]), reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)), dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=12, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.05, 0.05, 0.1, 0.1]), reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)), dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=12, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.033, 0.033, 0.067, 0.067]), reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)) ]), train_cfg=dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, match_low_quality=True, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=0, pos_weight=-1, debug=False), rpn_proposal=dict( nms_pre=2000, max_per_img=2000, nms=dict(type='nms', iou_threshold=0.7), min_bbox_size=0), rcnn=[ dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False), dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.6, neg_iou_thr=0.6, min_pos_iou=0.6, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False), dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.7, min_pos_iou=0.7, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False) ]), test_cfg=dict( rpn=dict( nms_pre=1000, max_per_img=1000, nms=dict(type='nms', iou_threshold=0.7), min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100))) dataset_type = 'CocoDataset' data_root = 'dataset/trainval/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) img_scale = (832, 832) train_pipeline = [ dict(type='Mosaic', img_scale=img_scale, pad_val=114.0), dict( type='RandomAffine', scaling_ratio_range=(0.1, 2), border=(-img_scale[0] // 2, -img_scale[1] // 2)), dict( type='MixUp', img_scale=img_scale, ratio_range=(0.8, 1.6), pad_val=114.0), dict( type='PhotoMetricDistortion', brightness_delta=32, contrast_range=(0.5, 1.5), saturation_range=(0.5, 1.5), hue_delta=18), dict(type='RandomFlip', flip_ratio=0.5), dict(type='Resize', keep_ratio=True), dict(type='Pad', pad_to_square=True, pad_val=114.0), dict(type='Normalize', **img_norm_cfg), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=img_scale, flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict(type='Pad', size=img_scale, pad_val=114.0), dict(type='Normalize', **img_norm_cfg), dict(type='ImageToTensor', keys=['img']), #try dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=8, workers_per_gpu=2, train=dict( type='MultiImageMixDataset', dataset=dict( type=dataset_type, ann_file= 'dataset/trainval/annotations/instances_train2017.json', img_prefix='dataset/trainval/', pipeline=[ dict(type='LoadImageFromFile', to_float32=True), dict(type='LoadAnnotations', with_bbox=True) ], filter_empty_gt=False, ), pipeline=train_pipeline, dynamic_scale=img_scale), val=dict( type=dataset_type, ann_file= 'dataset/trainval/annotations/instances_val2017.json', img_prefix='dataset/trainval/', pipeline=test_pipeline), test=dict( type=dataset_type, ann_file= 'dataset/trainval/annotations/instances_val2017.json', img_prefix='dataset/trainval/', ) ) evaluation = dict(interval=1, metric='bbox', save_best='bbox_mAP_50') optimizer = dict( type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005, nesterov=True, paramwise_cfg=dict(norm_decay_mult=0.0, bias_decay_mult=0.0)) optimizer_config = dict(grad_clip=None) lr_config = dict( policy='YOLOX', warmup='exp', by_epoch=False, warmup_by_epoch=True, warmup_ratio=1, warmup_iters=5, num_last_epochs=15, min_lr_ratio=0.05) runner = dict(type='EpochBasedRunner', max_epochs=40) checkpoint_config = dict(interval=1) log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')]) custom_hooks = [ dict(type='YOLOXModeSwitchHook', num_last_epochs=5, priority=48), dict( type='SyncRandomSizeHook', ratio_range=(14, 26), img_scale=(640, 640), interval=1, priority=48), dict(type='SyncNormHook', num_last_epochs=15, interval=1, priority=48), dict(type='ExpMomentumEMAHook', resume_from=None, priority=49) ] dist_params = dict(backend='nccl') log_level = 'INFO' load_from = None resume_from = None workflow = [('train', 1)] work_dir = './work_dirs/cascade_rcnn_x101_64x4d_fpn_20e_coco-3-832-mixupmosaic' gpu_ids = range(0, 1)

@jamiechoi1995 In fact, we did not test the above combination configuration, and there may be an incompatibility. > > > @jamiechoi1995 In fact, we did not test the above combination configuration, and there may be an incompatibility. I found that this error is due to the RandomAffine, MixUp and Mosaic augmentation returning Double type bbox, I solve it by forcing the bbox type of the above augmentation to float32. > > @jamiechoi1995 In fact, we did not test the above combination configuration, and there may be an incompatibility. > > I found that this error is due to the RandomAffine, MixUp and Mosaic augmentation returning Double type bbox, > I solve it by forcing the bbox type of the above augmentation to float32. Indeed possible. Can you create a PR to fix it? I find that it's because the dtype of warp_matrix in RandomAffine is float64, so after multiply this matrix to the box, the type of box becomes float64.

2021-08-23T10:42:16Z

[]

Traceback (most recent call last): File "tools/train.py", line 188, in <module> main() File "tools/train.py", line 184, in main meta=meta) File "mmdetection-2.15.1/mmdet/apis/train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], **kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 50, in train self.run_iter(data_batch, train_mode=True, **kwargs) File "mmcv-1.3.9/mmcv/runner/epoch_based_runner.py", line 30, in run_iter **kwargs) File "mmcv-1.3.9/mmcv/parallel/data_parallel.py", line 67, in train_step return self.module.train_step(*inputs[0], **kwargs[0]) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 237, in train_step losses = self(**data) File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__ result = self.forward(*input, **kwargs) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 98, in new_func return old_func(*args, **kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/base.py", line 171, in forward return self.forward_train(img, img_metas, **kwargs) File "mmdetection-2.15.1/mmdet/models/detectors/two_stage.py", line 140, in forward_train proposal_cfg=proposal_cfg) File "mmdetection-2.15.1/mmdet/models/dense_heads/base_dense_head.py", line 54, in forward_train losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) File "mmdetection-2.15.1/mmdet/models/dense_heads/rpn_head.py", line 74, in loss gt_bboxes_ignore=gt_bboxes_ignore) File "mmcv-1.3.9/mmcv/runner/fp16_utils.py", line 186, in new_func return old_func(*args, **kwargs) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 463, in loss label_channels=label_channels) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 345, in get_targets unmap_outputs=unmap_outputs) File "mmdetection-2.15.1/mmdet/core/utils/misc.py", line 29, in multi_apply return tuple(map(list, zip(*map_results))) File "mmdetection-2.15.1/mmdet/models/dense_heads/anchor_head.py", line 219, in _get_targets_single None if self.sampling else gt_labels) File "mmdetection-2.15.1/mmdet/core/bbox/assigners/max_iou_assigner.py", line 105, in assign overlaps = self.iou_calculator(gt_bboxes, bboxes) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 65, in __call__ return bbox_overlaps(bboxes1, bboxes2, mode, is_aligned) File "mmdetection-2.15.1/mmdet/core/bbox/iou_calculators/iou2d_calculator.py", line 233, in bbox_overlaps bboxes2[..., None, :, :2]) # [B, rows, cols, 2] RuntimeError: Expected object of scalar type Double but got scalar type Float for argument #2 'other' in call to _th_max

10,663

open-mmlab/mmdetection

open-mmlab__mmdetection-6079

b1f97c1b6d2b3c735d8c5690952264c598c2f206

diff --git a/mmdet/models/dense_heads/yolact_head.py b/mmdet/models/dense_heads/yolact_head.py --- a/mmdet/models/dense_heads/yolact_head.py +++ b/mmdet/models/dense_heads/yolact_head.py @@ -663,6 +663,10 @@ def _init_layers(self): protonets = protonets[:-1] return nn.Sequential(*protonets) + def forward_dummy(self, x): + prototypes = self.protonet(x) + return prototypes + def forward(self, x, coeff_pred, bboxes, img_meta, sampling_results=None): """Forward feature from the upstream network to get prototypes and linearly combine the prototypes, using masks coefficients, into diff --git a/mmdet/models/detectors/yolact.py b/mmdet/models/detectors/yolact.py --- a/mmdet/models/detectors/yolact.py +++ b/mmdet/models/detectors/yolact.py @@ -30,7 +30,10 @@ def forward_dummy(self, img): See `mmdetection/tools/analysis_tools/get_flops.py` """ - raise NotImplementedError + feat = self.extract_feat(img) + bbox_outs = self.bbox_head(feat) + prototypes = self.mask_head.forward_dummy(feat[0]) + return (bbox_outs, prototypes) def forward_train(self, img,

Issue about tools\analysis_tools\get_flops.py When I used the tools/analysis_tools/get_flops.py to compute the Flops and Parameters of YOLACT , I got an error "NotImplementedError". Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 722, in _call_impl result = self.forward(*input, **kwargs) File "/home/scsc01/lxx/mmlab/mmdet/models/detectors/yolact.py", line 32, in forward_dummy raise NotImplementedError NotImplementedError Does the get_flops.py not support the complexity calculation of YOLACT? thank you~~

In **yolact.py**, `forward_dummy` is not implemented https://github.com/open-mmlab/mmdetection/blob/6882fa0a220880e6e2e7e0536037f043b9031185/mmdet/models/detectors/yolact.py#L27-L32 But in **get_flops.py** https://github.com/open-mmlab/mmdetection/blob/6882fa0a220880e6e2e7e0536037f043b9031185/tools/analysis_tools/get_flops.py#L64-L69 I have checked `forward_dummy` of other models, and they are all implemented. I assume that the expected behavior is `forward_dummy` should not be declared in **yolact.py**, so the error information can be easier to understand? For now we have not supported it. We will support it in the future.

2021-09-10T13:27:44Z

[]

Traceback (most recent call last): File "tools/analysis_tools/get_flops.py", line 81, in <module> main() File "tools/analysis_tools/get_flops.py", line 71, in main flops, params = get_model_complexity_info(model, input_shape) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/mmcv/cnn/utils/flops_counter.py", line 104, in get_model_complexity_info _ = flops_model(batch) File "/home/scsc01/anaconda3/envs/mmlab/lib/python3.8/site-packages/torch/nn/modules/module.py", line 722, in _call_impl result = self.forward(*input, **kwargs) File "/home/scsc01/lxx/mmlab/mmdet/models/detectors/yolact.py", line 32, in forward_dummy raise NotImplementedError NotImplementedError

10,665

open-mmlab/mmdetection

open-mmlab__mmdetection-7147

ffff556edc5a96ce72ce5b4d27d1fbcaa0d22122

diff --git a/tools/analysis_tools/confusion_matrix.py b/tools/analysis_tools/confusion_matrix.py --- a/tools/analysis_tools/confusion_matrix.py +++ b/tools/analysis_tools/confusion_matrix.py @@ -207,7 +207,10 @@ def plot_confusion_matrix(confusion_matrix, ax.text( j, i, - '{}%'.format(int(confusion_matrix[i, j])), + '{}%'.format( + int(confusion_matrix[ + i, + j]) if not np.isnan(confusion_matrix[i, j]) else -1), ha='center', va='center', color='w',

Confusion matrix error While trying to generate the confusion matrix with this command: ``` python tools/analysis_tools/confusion_matrix.py ./work_dirs/perception-types--D06-01-2022--T09-23-45/perception-types.py results.pkl ./temp --show ``` I ran into this error: ``` Traceback (most recent call last): File "tools/analysis_tools/confusion_matrix.py", line 261, in <module> main() File "tools/analysis_tools/confusion_matrix.py", line 257, in main show=args.show) File "tools/analysis_tools/confusion_matrix.py", line 210, in plot_confusion_matrix '{}%'.format(int(confusion_matrix[i, j])), ValueError: cannot convert float NaN to integer ``` Would appreciate any help or suggestions! Thanks

I ran into the same problem so I would appreciate if someone could help us with this. Thanks for reporting the bug. We will fix it asap.

2022-02-12T15:36:58Z

[]

Traceback (most recent call last): File "tools/analysis_tools/confusion_matrix.py", line 261, in <module> main() File "tools/analysis_tools/confusion_matrix.py", line 257, in main show=args.show) File "tools/analysis_tools/confusion_matrix.py", line 210, in plot_confusion_matrix '{}%'.format(int(confusion_matrix[i, j])), ValueError: cannot convert float NaN to integer

10,676

open-mmlab/mmdetection

open-mmlab__mmdetection-7157

98949809b7179fab9391663ee5a4ab5978425f90

diff --git a/tools/deployment/onnx2tensorrt.py b/tools/deployment/onnx2tensorrt.py --- a/tools/deployment/onnx2tensorrt.py +++ b/tools/deployment/onnx2tensorrt.py @@ -201,7 +201,7 @@ def parse_args(): parsed directly from config file and are deprecated and will be \ removed in future releases.') if not args.input_img: - args.input_img = osp.join(osp.dirname(__file__), '../demo/demo.jpg') + args.input_img = osp.join(osp.dirname(__file__), '../../demo/demo.jpg') cfg = Config.fromfile(args.config)

FileNotFoundError: img file does not exist: tools/deployment/../demo/demo.jpg ``` Traceback (most recent call last): File "tools/deployment/onnx2tensorrt.py", line 255, in <module> verbose=args.verbose) File "tools/deployment/onnx2tensorrt.py", line 54, in onnx2tensorrt one_img, one_meta = preprocess_example_input(input_config) File "/workspace/mmdetection-2.18.1/mmdet/core/export/pytorch2onnx.py", line 139, in preprocess_example_input one_img = mmcv.imread(input_path) File "/workspace/mmcv-1.3.17-trt/mmcv/image/io.py", line 177, in imread f'img file does not exist: {img_or_path}') File "/workspace/mmcv-1.3.17-trt/mmcv/utils/path.py", line 23, in check_file_exist raise FileNotFoundError(msg_tmpl.format(filename)) FileNotFoundError: img file does not exist: tools/deployment/../demo/demo.jpg ```

2022-02-14T12:37:02Z

[]

Traceback (most recent call last): File "tools/deployment/onnx2tensorrt.py", line 255, in <module> verbose=args.verbose) File "tools/deployment/onnx2tensorrt.py", line 54, in onnx2tensorrt one_img, one_meta = preprocess_example_input(input_config) File "/workspace/mmdetection-2.18.1/mmdet/core/export/pytorch2onnx.py", line 139, in preprocess_example_input one_img = mmcv.imread(input_path) File "/workspace/mmcv-1.3.17-trt/mmcv/image/io.py", line 177, in imread f'img file does not exist: {img_or_path}') File "/workspace/mmcv-1.3.17-trt/mmcv/utils/path.py", line 23, in check_file_exist raise FileNotFoundError(msg_tmpl.format(filename)) FileNotFoundError: img file does not exist: tools/deployment/../demo/demo.jpg

10,677

open-mmlab/mmdetection

open-mmlab__mmdetection-7407

c546b5044098b71d59a139036a87c5c97bcab4e2

diff --git a/tools/analysis_tools/analyze_logs.py b/tools/analysis_tools/analyze_logs.py old mode 100644 new mode 100755 --- a/tools/analysis_tools/analyze_logs.py +++ b/tools/analysis_tools/analyze_logs.py @@ -17,6 +17,10 @@ def cal_train_time(log_dicts, args): all_times.append(log_dict[epoch]['time']) else: all_times.append(log_dict[epoch]['time'][1:]) + if not all_times: + raise KeyError( + 'Please reduce the log interval in the config so that' + 'interval is less than iterations of one epoch.') all_times = np.array(all_times) epoch_ave_time = all_times.mean(-1) slowest_epoch = epoch_ave_time.argmax() @@ -50,12 +54,21 @@ def plot_curve(log_dicts, args): epochs = list(log_dict.keys()) for j, metric in enumerate(metrics): print(f'plot curve of {args.json_logs[i]}, metric is {metric}') - if metric not in log_dict[epochs[0]]: + if metric not in log_dict[epochs[int(args.start_epoch) - 1]]: + if 'mAP' in metric: + raise KeyError( + f'{args.json_logs[i]} does not contain metric ' + f'{metric}. Please check if "--no-validate" is ' + 'specified when you trained the model.') raise KeyError( - f'{args.json_logs[i]} does not contain metric {metric}') + f'{args.json_logs[i]} does not contain metric {metric}. ' + 'Please reduce the log interval in the config so that ' + 'interval is less than iterations of one epoch.') if 'mAP' in metric: - xs = np.arange(1, max(epochs) + 1) + xs = np.arange( + int(args.start_epoch), + max(epochs) + 1, int(args.eval_interval)) ys = [] for epoch in epochs: ys += log_dict[epoch][metric] @@ -104,6 +117,16 @@ def add_plot_parser(subparsers): nargs='+', default=['bbox_mAP'], help='the metric that you want to plot') + parser_plt.add_argument( + '--start-epoch', + type=str, + default='1', + help='the epoch that you want to start') + parser_plt.add_argument( + '--eval-interval', + type=str, + default='1', + help='the eval interval when training') parser_plt.add_argument('--title', type=str, help='title of figure') parser_plt.add_argument( '--legend',

./tools/analysis_tools/analyze_logs.py plot_curve IndexError: list index out of range `(openmmlab) lbc@prust-System-3:~/mmdetection-master$ python3.8 ./tools/analysis_tools/analyze_logs.py plot_curve ./work_dirs/deformable_detr_twostage_refine_r50_16x2_50e_coco/20211119_170702.log.json --keys bbox_mAP plot curve of ./work_dirs/deformable_detr_twostage_refine_r50_16x2_50e_coco/20211119_170702.log.json, metric is bbox_mAP Traceback (most recent call last): File "./tools/analysis_tools/analyze_logs.py", line 180, in <module> main() File "./tools/analysis_tools/analyze_logs.py", line 176, in main eval(args.task)(log_dicts, args) File "./tools/analysis_tools/analyze_logs.py", line 53, in plot_curve if metric not in log_dict[epochs[0]]: IndexError: list index out of range `

2022-03-15T10:49:28Z

[]

Traceback (most recent call last): File "./tools/analysis_tools/analyze_logs.py", line 180, in <module> main() File "./tools/analysis_tools/analyze_logs.py", line 176, in main eval(args.task)(log_dicts, args) File "./tools/analysis_tools/analyze_logs.py", line 53, in plot_curve if metric not in log_dict[epochs[0]]: IndexError: list index out of range

10,680

open-mmlab/mmdetection

open-mmlab__mmdetection-8273

ca11860f4f3c3ca2ce8340e2686eeaec05b29111

diff --git a/mmdet/core/hook/wandblogger_hook.py b/mmdet/core/hook/wandblogger_hook.py --- a/mmdet/core/hook/wandblogger_hook.py +++ b/mmdet/core/hook/wandblogger_hook.py @@ -135,7 +135,8 @@ def before_run(self, runner): super(MMDetWandbHook, self).before_run(runner) # Save and Log config. - if runner.meta is not None: + if runner.meta is not None and runner.meta.get('exp_name', + None) is not None: src_cfg_path = osp.join(runner.work_dir, runner.meta.get('exp_name', None)) if osp.exists(src_cfg_path):

WandbLogger Hook Error WandbLogger Hook error Code ``` log_config = dict( interval=10, hooks=[ dict(type='TensorboardLoggerHook'), dict(type='TextLoggerHook'), dict(type='MMDetWandbHook', init_kwargs={ 'project': PROJECT, 'entity': ENTITY, 'name': TAG, 'config': { 'lr': 0.0025, 'batch_size':16 }, 'tags': WANDB_TAGS }, interval=1, log_checkpoint=False, log_checkpoint_metadata=True, num_eval_images=3, bbox_score_thr=0.3 ) ]) ``` **Environment** ``` Python: 3.7.11 (default, Jul 27 2021, 14:32:16) [GCC 7.5.0] CUDA available: True GPU 0: Tesla V100-SXM2-16GB CUDA_HOME: /usr/local/cuda-10.1 NVCC: Cuda compilation tools, release 10.1, V10.1.24 GCC: gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 PyTorch: 1.10.0+cu102 PyTorch compiling details: PyTorch built with: - GCC 7.3 - C++ Version: 201402 - Intel(R) oneAPI Math Kernel Library Version 2021.4-Product Build 20210904 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.2.3 (Git Hash 7336ca9f055cf1bfa13efb658fe15dc9b41f0740) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70 - CuDNN 7.6.5 - Magma 2.5.2 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=10.2, CUDNN_VERSION=7.6.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.10.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, TorchVision: 0.11.1+cu102 OpenCV: 4.5.4 MMCV: 1.5.3 MMCV Compiler: GCC 7.3 MMCV CUDA Compiler: 10.2 MMDetection: 2.25.0+55d536e ``` **Error traceback** ``` Traceback (most recent call last): File "train.py", line 28, in <module> fire.Fire(launch) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 141, in Fire component_trace = _Fire(component, args, parsed_flag_args, context, name) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 471, in _Fire target=component.__name__) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 681, in _CallAndUpdateTrace component = fn(*varargs, **kwargs) File "train.py", line 25, in launch train_detector(model, datasets, cfg, distributed=False, validate=True) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/apis/train.py", line 244, in train_detector runner.run(data_loaders, cfg.workflow) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 111, in run self.call_hook('before_run') File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/base_runner.py", line 309, in call_hook getattr(hook, fn_name)(self) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/dist_utils.py", line 135, in wrapper return func(*args, **kwargs) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/core/hook/wandblogger_hook.py", line 140, in before_run runner.meta.get('exp_name', None)) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/posixpath.py", line 94, in join genericpath._check_arg_types('join', a, *p) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/genericpath.py", line 153, in _check_arg_types (funcname, s.__class__.__name__)) from None TypeError: join() argument must be str or bytes, not 'NoneType' ```

Hey @sarmientoj24 this seems like a quick fix. ``` # Save and Log config. if runner.meta is not None and runner.meta.get('exp_name', None) is not None: src_cfg_path = osp.join(runner.work_dir, runner.meta.get('exp_name', None)) if osp.exists(src_cfg_path): self.wandb.save(src_cfg_path, base_path=runner.work_dir) self._update_wandb_config(runner) else: runner.logger.warning('No meta information found in the runner. ') ``` I will make a PR to fix this.

2022-06-27T11:30:01Z

[]

Traceback (most recent call last): File "train.py", line 28, in <module> fire.Fire(launch) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 141, in Fire component_trace = _Fire(component, args, parsed_flag_args, context, name) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 471, in _Fire target=component.__name__) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/fire/core.py", line 681, in _CallAndUpdateTrace component = fn(*varargs, **kwargs) File "train.py", line 25, in launch train_detector(model, datasets, cfg, distributed=False, validate=True) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/apis/train.py", line 244, in train_detector runner.run(data_loaders, cfg.workflow) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 111, in run self.call_hook('before_run') File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/base_runner.py", line 309, in call_hook getattr(hook, fn_name)(self) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/site-packages/mmcv/runner/dist_utils.py", line 135, in wrapper return func(*args, **kwargs) File "/home/james_sarmiento/mmdetection/mmdetection/mmdet/core/hook/wandblogger_hook.py", line 140, in before_run runner.meta.get('exp_name', None)) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/posixpath.py", line 94, in join genericpath._check_arg_types('join', a, *p) File "/home/james_sarmiento/anaconda3/envs/yl/lib/python3.7/genericpath.py", line 153, in _check_arg_types (funcname, s.__class__.__name__)) from None TypeError: join() argument must be str or bytes, not 'NoneType'

10,692

open-mmlab/mmdetection

open-mmlab__mmdetection-854

ae856e11ec3f281ce77c3f8fdf1cd87162598ffb

diff --git a/mmdet/models/backbones/resnet.py b/mmdet/models/backbones/resnet.py --- a/mmdet/models/backbones/resnet.py +++ b/mmdet/models/backbones/resnet.py @@ -297,11 +297,11 @@ def make_res_layer(block, layers = [] layers.append( block( - inplanes, - planes, - stride, - dilation, - downsample, + inplanes=inplanes, + planes=planes, + stride=stride, + dilation=dilation, + downsample=downsample, style=style, with_cp=with_cp, conv_cfg=conv_cfg, @@ -314,10 +314,10 @@ def make_res_layer(block, for i in range(1, blocks): layers.append( block( - inplanes, - planes, - 1, - dilation, + inplanes=inplanes, + planes=planes, + stride=1, + dilation=dilation, style=style, with_cp=with_cp, conv_cfg=conv_cfg, diff --git a/mmdet/models/backbones/resnext.py b/mmdet/models/backbones/resnext.py --- a/mmdet/models/backbones/resnext.py +++ b/mmdet/models/backbones/resnext.py @@ -11,12 +11,12 @@ class Bottleneck(_Bottleneck): - def __init__(self, groups=1, base_width=4, *args, **kwargs): + def __init__(self, inplanes, planes, groups=1, base_width=4, **kwargs): """Bottleneck block for ResNeXt. If style is "pytorch", the stride-two layer is the 3x3 conv layer, if it is "caffe", the stride-two layer is the first 1x1 conv layer. """ - super(Bottleneck, self).__init__(*args, **kwargs) + super(Bottleneck, self).__init__(inplanes, planes, **kwargs) if groups == 1: width = self.planes

GCNNet with x101 backbone is not work I use gcb in x101 backbone，but get this error: Traceback (most recent call last): File "./tools/train.py", line 95, in <module> main() File "./tools/train.py", line 73, in main cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 60, in build_detector return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(**args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/detectors/cascade_rcnn.py", line 36, in __init__ self.backbone = builder.build_backbone(backbone) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 36, in build_backbone return build(cfg, BACKBONES) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(**args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 191, in __init__ super(ResNeXt, self).__init__(**kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 447, in __init__ gen_attention_blocks=stage_with_gen_attention[i]) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 317, in make_res_layer (0 in gen_attention_blocks) else None)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 19, in __init__ super(Bottleneck, self).__init__(*args, **kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 205, in __init__ self.context_block = ContextBlock(inplanes=gcb_inplanes, **gcb) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/ops/gcb/context_block.py", line 40, in __init__ nn.Conv2d(self.inplanes, self.planes, kernel_size=1), File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 327, in __init__ False, _pair(0), groups, bias, padding_mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 45, in __init__ self.reset_parameters() File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 48, in reset_parameters init.kaiming_uniform_(self.weight, a=math.sqrt(5)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 323, in kaiming_uniform_ fan = _calculate_correct_fan(tensor, mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 292, in _calculate_correct_fan fan_in, fan_out = _calculate_fan_in_and_fan_out(tensor) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 222, in _calculate_fan_in_and_fan_out receptive_field_size = tensor[0][0].numel() IndexError: index 0 is out of bounds for dimension 0 with size 0

i meet the same problem. Do you know what's wrong? you should update your backbone file，because the orfer of parameters is changed to compliance with specificatinos. See pull #780 Thanks for reporting! Will fix soon. @zhengye1995 i don't know what's your mean? your mean is my hyper parameter have some wrong? but when i run r4,it is ok, i only change gcb 1/4 to 1/16, the problem will appear my backbone as following: backbone=dict( type='ResNeXt', depth=101, groups=32, base_width=4, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, style='pytorch', gcb = dict( ratio=1. / 16., ), stage_with_gcb = (False, True, True, True), dcn=dict( modulated=False, groups=32, deformable_groups=1, fallback_on_stride=False), stage_with_dcn=(False, True, True, True), norm_cfg=norm_cfg, norm_eval=False, ),

2019-06-22T15:25:21Z

[]

Traceback (most recent call last): File "./tools/train.py", line 95, in <module> main() File "./tools/train.py", line 73, in main cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 60, in build_detector return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(**args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/detectors/cascade_rcnn.py", line 36, in __init__ self.backbone = builder.build_backbone(backbone) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 36, in build_backbone return build(cfg, BACKBONES) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 32, in build return _build_module(cfg, registry, default_args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/builder.py", line 24, in _build_module return obj_type(**args) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 191, in __init__ super(ResNeXt, self).__init__(**kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 447, in __init__ gen_attention_blocks=stage_with_gen_attention[i]) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 317, in make_res_layer (0 in gen_attention_blocks) else None)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnext.py", line 19, in __init__ super(Bottleneck, self).__init__(*args, **kwargs) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/models/backbones/resnet.py", line 205, in __init__ self.context_block = ContextBlock(inplanes=gcb_inplanes, **gcb) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/mmdet-0.6.0+unknown-py3.6.egg/mmdet/ops/gcb/context_block.py", line 40, in __init__ nn.Conv2d(self.inplanes, self.planes, kernel_size=1), File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 327, in __init__ False, _pair(0), groups, bias, padding_mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 45, in __init__ self.reset_parameters() File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 48, in reset_parameters init.kaiming_uniform_(self.weight, a=math.sqrt(5)) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 323, in kaiming_uniform_ fan = _calculate_correct_fan(tensor, mode) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 292, in _calculate_correct_fan fan_in, fan_out = _calculate_fan_in_and_fan_out(tensor) File "/home/zhengye/anaconda3/envs/pytorch1.0/lib/python3.6/site-packages/torch/nn/init.py", line 222, in _calculate_fan_in_and_fan_out receptive_field_size = tensor[0][0].numel() IndexError: index 0 is out of bounds for dimension 0 with size 0

10,694

open-mmlab/mmdetection

open-mmlab__mmdetection-9151

e71b499608e9c3ccd4211e7c815fa20eeedf18a2

diff --git a/mmdet/models/detectors/rpn.py b/mmdet/models/detectors/rpn.py --- a/mmdet/models/detectors/rpn.py +++ b/mmdet/models/detectors/rpn.py @@ -1,5 +1,6 @@ # Copyright (c) OpenMMLab. All rights reserved. import warnings +from inspect import signature import mmcv import torch @@ -153,7 +154,9 @@ def show_result(self, data, result, top_k=20, **kwargs): np.ndarray: The image with bboxes drawn on it. """ if kwargs is not None: - kwargs.pop('score_thr', None) - kwargs.pop('text_color', None) - kwargs['colors'] = kwargs.pop('bbox_color', 'green') + kwargs['colors'] = 'green' + sig = signature(mmcv.imshow_bboxes) + for k in list(kwargs.keys()): + if k not in sig.parameters: + kwargs.pop(k) mmcv.imshow_bboxes(data, result, top_k=top_k, **kwargs)

run image_demo.py with cascade_rpn model show error ### Prerequisite - [X] I have searched [the existing and past issues](https://github.com/open-mmlab/mmdetection/issues) but cannot get the expected help. - [X] I have read the [FAQ documentation](https://mmdetection.readthedocs.io/en/latest/faq.html) but cannot get the expected help. - [X] The bug has not been fixed in the [latest version](https://github.com/open-mmlab/mmdetection). ### 🐞 Describe the bug ``` python demo/image_demo.py demo/demo.jpg configs/cascade_rpn/crpn_r50_caffe_fpn_1x_coco.py checkpoints/cascade_rpn_r50_caffe_fpn_1x_coco-7aa93cef.pth``` ``` UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:2895.) return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined] Traceback (most recent call last): File "demo/image_demo.py", line 68, in <module> main(args) File "demo/image_demo.py", line 38, in main show_result_pyplot( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/apis/inference.py", line 241, in show_result_pyplot model.show_result( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/models/detectors/rpn.py", line 159, in show_result mmcv.imshow_bboxes(data, result, top_k=top_k, **kwargs) TypeError: imshow_bboxes() got an unexpected keyword argument 'mask_color' ``` ### Environment ``` Python: 3.8.13 (default, Mar 28 2022, 11:38:47) [GCC 7.5.0] CUDA available: True GPU 0: GeForce RTX 2080 SUPER CUDA_HOME: /usr/local/cuda-10.2 NVCC: Cuda compilation tools, release 10.2, V10.2.8 GCC: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609 PyTorch: 1.12.1+cu102 PyTorch compiling details: PyTorch built with: - GCC 7.3 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 10.2 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70 - CuDNN 7.6.5 - Magma 2.5.2 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=10.2, CUDNN_VERSION=7.6.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.12.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=OFF, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, TorchVision: 0.13.1+cu102 OpenCV: 4.6.0 MMCV: 1.6.2 MMCV Compiler: GCC 7.3 MMCV CUDA Compiler: 10.2 MMDetection: 2.25.2+9d3e162 ``` ### Additional information _No response_

Thanks for your report! We will fix this bug in the next version.

2022-10-27T03:49:51Z

[]

Traceback (most recent call last): File "demo/image_demo.py", line 68, in <module> main(args) File "demo/image_demo.py", line 38, in main show_result_pyplot( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/apis/inference.py", line 241, in show_result_pyplot model.show_result( File "/home/ngi/IdeaProjects/mmlab/mmdetection/mmdet/models/detectors/rpn.py", line 159, in show_result mmcv.imshow_bboxes(data, result, top_k=top_k, **kwargs) TypeError: imshow_bboxes() got an unexpected keyword argument 'mask_color'

10,695

open-mmlab/mmdetection

open-mmlab__mmdetection-9694

ea29a1edbe973389c2d705d99fd060b07decddab

diff --git a/mmdet/utils/__init__.py b/mmdet/utils/__init__.py --- a/mmdet/utils/__init__.py +++ b/mmdet/utils/__init__.py @@ -7,6 +7,7 @@ from .memory import AvoidCUDAOOM, AvoidOOM from .misc import find_latest_checkpoint, update_data_root from .replace_cfg_vals import replace_cfg_vals +from .rfnext import rfnext_init_model from .setup_env import setup_multi_processes from .split_batch import split_batch from .util_distribution import build_ddp, build_dp, get_device @@ -16,5 +17,6 @@ 'update_data_root', 'setup_multi_processes', 'get_caller_name', 'log_img_scale', 'compat_cfg', 'split_batch', 'build_ddp', 'build_dp', 'get_device', 'replace_cfg_vals', 'AvoidOOM', 'AvoidCUDAOOM', - 'get_max_num_gt_division_factor', 'masked_fill', 'batch_images_to_levels' + 'get_max_num_gt_division_factor', 'masked_fill', 'batch_images_to_levels', + 'rfnext_init_model' ]

[Bug] Training Error ### Prerequisite - [X] I have searched [Issues](https://github.com/open-mmlab/mmdetection/issues) and [Discussions](https://github.com/open-mmlab/mmdetection/discussions) but cannot get the expected help. - [X] I have read the [FAQ documentation](https://mmdetection.readthedocs.io/en/latest/faq.html) but cannot get the expected help. - [X] The bug has not been fixed in the [latest version (master)](https://github.com/open-mmlab/mmdetection) or [latest version (3.x)](https://github.com/open-mmlab/mmdetection/tree/dev-3.x). ### Task I have modified the scripts/configs, or I'm working on my own tasks/models/datasets. ### Branch master branch https://github.com/open-mmlab/mmdetection ### Environment ``` sys.platform: linux Python: 3.7.15 (default, Nov 24 2022, 21:12:53) [GCC 11.2.0] CUDA available: True GPU 0,1,2,3,4,5,6,7: NVIDIA GeForce RTX 3090 CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 11.6, V11.6.55 GCC: gcc (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0 PyTorch: 1.13.1+cu117 PyTorch compiling details: PyTorch built with: - GCC 9.3 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 11.7 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86 - CuDNN 8.5 - Magma 2.6.1 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.13.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, TorchVision: 0.14.1+cu117 OpenCV: 4.7.0 MMCV: 1.7.1 MMCV Compiler: GCC 9.3 MMCV CUDA Compiler: 11.7 MMDetection: 2.28.0+b955832 ``` ### Reproduces the problem - code sample ```python python train.py $CONFIG_FILE ``` ### Reproduces the problem - command or script ```python python train.py $CONFIG_FILE ``` ### Reproduces the problem - error message ``` /anaconda3/envs/dl/lib/python3.7/site-packages/mmcv/__init__.py:21: UserWarning: On January 1, 2023, MMCV will release v2.0.0, in which it will remove components related to the training process and add a data transformation module. In addition, it will rename the package names mmcv to mmcv-lite and mmcv-full to mmcv. See https://github.com/open-mmlab/mmcv/blob/master/docs/en/compatibility.md for more details. 'On January 1, 2023, MMCV will release v2.0.0, in which it will remove ' Traceback (most recent call last): File "train.py", line 20, in <module> from mmdet.utils import (collect_env, get_device, get_root_logger, ImportError: cannot import name 'rfnext_init_model' from 'mmdet.utils' (/mmdetection/mmdet/utils/__init__.py) ``` ### Additional information I clone the latest mmdetection package and intend to run a couple of benchmark. However, neither my configs nor the configs provided in the `mmdet/configs` yields the abovementioned error message.

I also encountered this problem

2023-01-30T08:09:21Z

[]

Traceback (most recent call last): File "train.py", line 20, in <module> from mmdet.utils import (collect_env, get_device, get_root_logger, ImportError: cannot import name 'rfnext_init_model' from 'mmdet.utils' (/mmdetection/mmdet/utils/__init__.py)

10,701

pandas-dev/pandas

pandas-dev__pandas-10108

eafd22d961934a7b3cc72607ef4512a18b419085

diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -57,3 +57,5 @@ Performance Improvements Bug Fixes ~~~~~~~~~ + +- Bug in ``Categorical`` repr with ``display.width`` of ``None`` in Python 3 (:issue:`10087`) diff --git a/pandas/core/categorical.py b/pandas/core/categorical.py --- a/pandas/core/categorical.py +++ b/pandas/core/categorical.py @@ -1310,8 +1310,7 @@ def _repr_categories_info(self): levheader = "Categories (%d, %s): " % (len(self.categories), self.categories.dtype) width, height = get_terminal_size() - max_width = (width if get_option("display.width") == 0 - else get_option("display.width")) + max_width = get_option("display.width") or width if com.in_ipython_frontend(): # 0 = no breaks max_width = 0

BUG: categorical doesn't handle display.width of None in Python 3 Categorical Series have a special repr that looks at display.width, which can be None if following the Options and Settings docs. Unlike Python 2, in Python 3 an integer vs None comparison throws an exception. (on current master, and has been true for several releases now) ``` python Python 3.4.0 (default, Apr 11 2014, 13:05:11) [GCC 4.8.2] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import pandas as pd >>> pd.core.config.set_option('display.width', None) >>> import numpy as np >>> x = pd.Series(np.random.randn(100)) >>> pd.cut(x, 10) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 67, in __repr__ return str(self) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 46, in __str__ return self.__unicode__() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 897, in __unicode__ max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 962, in to_string name=name, max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 992, in _get_repr result = formatter.to_string() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 222, in to_string footer = self._get_footer() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 196, in _get_footer level_info = self.tr_series.values._repr_categories_info() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/categorical.py", line 1323, in _repr_categories_info if max_width != 0 and cur_col_len + sep_len + len(val) > max_width: TypeError: unorderable types: int() > NoneType() ```

should be a simple fix. PR's are welcome. Sure, I'll try and get to it soon.

2015-05-11T22:15:45Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 67, in __repr__ return str(self) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/base.py", line 46, in __str__ return self.__unicode__() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 897, in __unicode__ max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 962, in to_string name=name, max_rows=max_rows) File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/series.py", line 992, in _get_repr result = formatter.to_string() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 222, in to_string footer = self._get_footer() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/format.py", line 196, in _get_footer level_info = self.tr_series.values._repr_categories_info() File "/home/andrew/git/pandas-rosnfeld-py3/pandas/core/categorical.py", line 1323, in _repr_categories_info if max_width != 0 and cur_col_len + sep_len + len(val) > max_width: TypeError: unorderable types: int() > NoneType()

10,720

pandas-dev/pandas

pandas-dev__pandas-10473

83b232089b9292b11d4b9b00c0e50cc4a829f016

diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -27,6 +27,7 @@ New features ~~~~~~~~~~~~ - SQL io functions now accept a SQLAlchemy connectable. (:issue:`7877`) +- Enable writing complex values to HDF stores when using table format (:issue:`10447`) .. _whatsnew_0170.enhancements.other: @@ -147,3 +148,4 @@ Bug Fixes - Bug in `groupby.var` which caused variance to be inaccurate for small float values (:issue:`10448`) - Bug in ``Series.plot(kind='hist')`` Y Label not informative (:issue:`10485`) + diff --git a/pandas/io/pytables.py b/pandas/io/pytables.py --- a/pandas/io/pytables.py +++ b/pandas/io/pytables.py @@ -1773,6 +1773,8 @@ def set_kind(self): self.kind = 'string' elif dtype.startswith(u('float')): self.kind = 'float' + elif dtype.startswith(u('complex')): + self.kind = 'complex' elif dtype.startswith(u('int')) or dtype.startswith(u('uint')): self.kind = 'integer' elif dtype.startswith(u('date')): @@ -1802,6 +1804,8 @@ def set_atom(self, block, block_items, existing_col, min_itemsize, return self.set_atom_datetime64(block) elif block.is_timedelta: return self.set_atom_timedelta64(block) + elif block.is_complex: + return self.set_atom_complex(block) dtype = block.dtype.name inferred_type = lib.infer_dtype(block.values) @@ -1936,6 +1940,12 @@ def get_atom_coltype(self, kind=None): def get_atom_data(self, block, kind=None): return self.get_atom_coltype(kind=kind)(shape=block.shape[0]) + def set_atom_complex(self, block): + self.kind = block.dtype.name + itemsize = int(self.kind.split('complex')[-1]) // 8 + self.typ = _tables().ComplexCol(itemsize=itemsize, shape=block.shape[0]) + self.set_data(block.values.astype(self.typ.type, copy=False)) + def set_atom_data(self, block): self.kind = block.dtype.name self.typ = self.get_atom_data(block) @@ -3147,8 +3157,8 @@ def f(i, c): def create_index(self, columns=None, optlevel=None, kind=None): """ Create a pytables index on the specified columns - note: cannot index Time64Col() currently; PyTables must be >= 2.3 - + note: cannot index Time64Col() or ComplexCol currently; + PyTables must be >= 3.0 Paramaters ---------- @@ -3203,6 +3213,12 @@ def create_index(self, columns=None, optlevel=None, kind=None): # create the index if not v.is_indexed: + if v.type.startswith('complex'): + raise TypeError('Columns containing complex values can be stored but cannot' + ' be indexed when using table format. Either use fixed ' + 'format, set index=False, or do not include the columns ' + 'containing complex values to data_columns when ' + 'initializing the table.') v.create_index(**kw) def read_axes(self, where, **kwargs):

BUG: Cannot store complex valued Series/DataFrame/Panel/Panel4D as 'table' to hdf Attempting to store a Panel4D of complext128 results in an error. This is true if I use table or fixed formats (different errors) [I am showing the last call from pandas as well as the actual error] ``` python pd.Panel4D(np.tile(np.array([1+1j,2+2j]),[2,2,2,1])).to_hdf('complex.h5','complex',format='t') Traceback (most recent call last): ... File "/miniconda/envs/py34/lib/python3.4/site-packages/pandas/io/pytables.py", line 1930, in get_atom_data return self.get_atom_coltype(kind=kind)(shape=block.shape[0]) ... File "/miniconda/envs/py34/lib/python3.4/site-packages/tables/atom.py", line 740, in __init__ "to avoid confusions with PyTables 1.X complex atom names, " TypeError: to avoid confusions with PyTables 1.X complex atom names, please use ``ComplexAtom(itemsize=N)``, where N=8 for single precision complex atoms, and N=16 for double precision complex atoms ``` ``` INSTALLED VERSIONS ------------------ commit: None python: 3.4.3.final.0 python-bits: 64 OS: Linux OS-release: 2.6.32-504.16.2.el6.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 pandas: 0.16.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: None statsmodels: None IPython: 3.1.0 sphinx: None patsy: 0.3.0 dateutil: 2.4.2 pytz: 2015.4 bottleneck: None tables: 3.1.1 numexpr: 2.3.1 matplotlib: None openpyxl: 2.0.2 xlrd: 0.9.3 xlwt: 1.0.0 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None ``` If you try fixed, you get ``` TypeError: cannot properly create the storer for: [_STORER_MAP] [group->/complex (Group) '',value-><class 'pandas.core.panelnd.Panel4D'>,format->fixed,append->False,kwargs->{'encoding': None}] ```

'>3d' not supported in fixed at all (as pretty useless - u generally need/want to append to these) complex is not supported very well in the exporters but of course fixes welcome Turns out this is true for all types, not just Panel4D

2015-06-29T19:27:56Z

[]

Traceback (most recent call last): ... File "/miniconda/envs/py34/lib/python3.4/site-packages/pandas/io/pytables.py", line 1930, in get_atom_data return self.get_atom_coltype(kind=kind)(shape=block.shape[0]) ... File "/miniconda/envs/py34/lib/python3.4/site-packages/tables/atom.py", line 740, in __init__ "to avoid confusions with PyTables 1.X complex atom names, " TypeError: to avoid confusions with PyTables 1.X complex atom names, please use ``ComplexAtom(itemsize=N)``, where N=8 for single precision complex atoms, and N=16 for double precision complex atoms

10,768

pandas-dev/pandas

pandas-dev__pandas-10497

bbec57d6f881cb7d26ad65319595c8594381fe8c

diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -128,6 +128,7 @@ Bug Fixes - Bug in ``test_categorical`` on big-endian builds (:issue:`10425`) +- Bug in ``Series.shift`` and ``DataFrame.shift`` not supporting categorical data (:issue:`9416`) - Bug in ``Series.map`` using categorical ``Series`` raises ``AttributeError`` (:issue:`10324`) - Bug in ``MultiIndex.get_level_values`` including ``Categorical`` raises ``AttributeError`` (:issue:`10460`) diff --git a/pandas/core/categorical.py b/pandas/core/categorical.py --- a/pandas/core/categorical.py +++ b/pandas/core/categorical.py @@ -820,6 +820,35 @@ def shape(self): return tuple([len(self._codes)]) + def shift(self, periods): + """ + Shift Categorical by desired number of periods. + + Parameters + ---------- + periods : int + Number of periods to move, can be positive or negative + + Returns + ------- + shifted : Categorical + """ + # since categoricals always have ndim == 1, an axis parameter + # doesnt make any sense here. + codes = self.codes + if codes.ndim > 1: + raise NotImplementedError("Categorical with ndim > 1.") + if np.prod(codes.shape) and (periods != 0): + codes = np.roll(codes, com._ensure_platform_int(periods), axis=0) + if periods > 0: + codes[:periods] = -1 + else: + codes[periods:] = -1 + + return Categorical.from_codes(codes, + categories=self.categories, + ordered=self.ordered) + def __array__(self, dtype=None): """ The numpy array interface. diff --git a/pandas/core/internals.py b/pandas/core/internals.py --- a/pandas/core/internals.py +++ b/pandas/core/internals.py @@ -1709,6 +1709,10 @@ def interpolate(self, method='pad', axis=0, inplace=False, limit=limit), placement=self.mgr_locs) + def shift(self, periods, axis=0): + return self.make_block_same_class(values=self.values.shift(periods), + placement=self.mgr_locs) + def take_nd(self, indexer, axis=0, new_mgr_locs=None, fill_tuple=None): """ Take values according to indexer and return them as a block.bb

Series.shift() doesn't work for categorical type Not sure if this is intentional, but Series.shift() won't run with categorical dtypes: ``` python ser = pd.Series(['a', 'b', 'c', 'd'], dtype="category") ser.shift(1) Traceback (most recent call last): File "<ipython-input-15-1a7536b0af06>", line 1, in <module> ser.shift(1) File "/.../pandas/core/generic.py", line 3394, in shift new_data = self._data.shift(periods=periods, axis=block_axis) File "/.../pandas/core/internals.py", line 2533, in shift return self.apply('shift', **kwargs) File "/.../pandas/core/internals.py", line 2497, in apply applied = getattr(b, f)(**kwargs) File "/.../pandas/core/internals.py", line 893, in shift new_values, fill_value = com._maybe_upcast(self.values) File "/.../pandas/core/common.py", line 1218, in _maybe_upcast new_dtype, fill_value = _maybe_promote(dtype, fill_value) File "/.../pandas/core/common.py", line 1124, in _maybe_promote if issubclass(np.dtype(dtype).type, compat.string_types): TypeError: data type not understood ```

This simply hasn't been implemented, but otherwise was not intentional. Help would be appreciated if you're interested in putting together a PR. The place to get started (I believe) would be to implement the `shift` method on `CategoricalBlock` in `pandas.core.internals`. here's basically what you would do ``` In [2]: s = Series(list('aabbcde'),dtype='category') In [3]: s Out[3]: 0 a 1 a 2 b 3 b 4 c 5 d 6 e dtype: category Categories (5, object): [a < b < c < d < e] In [4]: s.values Out[4]: [a, a, b, b, c, d, e] Categories (5, object): [a < b < c < d < e] In [5]: s.values.codes Out[5]: array([0, 0, 1, 1, 2, 3, 4], dtype=int8) In [6]: np.roll(s.values.codes,len(s)-1,axis=0) Out[6]: array([0, 1, 1, 2, 3, 4, 0], dtype=int8) In [7]: codes = np.roll(s.values.codes,len(s)-1,axis=0) In [8]: codes[-1] = -1 In [11]: pd.Categorical(codes,categories=s.values.categories,fastpath=True) Out[11]: [a, b, b, c, d, e, NaN] Categories (5, object): [a, b, c, d, e] ``` you would use the Block.shift method (and pass the codes to it for the actual shifting), then wrap it back to a catetgorical (their is a method for that too). Should be pretty straightforward.

2015-07-03T13:42:01Z

[]

Traceback (most recent call last): File "<ipython-input-15-1a7536b0af06>", line 1, in <module> ser.shift(1) File "/.../pandas/core/generic.py", line 3394, in shift new_data = self._data.shift(periods=periods, axis=block_axis) File "/.../pandas/core/internals.py", line 2533, in shift return self.apply('shift', **kwargs) File "/.../pandas/core/internals.py", line 2497, in apply applied = getattr(b, f)(**kwargs) File "/.../pandas/core/internals.py", line 893, in shift new_values, fill_value = com._maybe_upcast(self.values) File "/.../pandas/core/common.py", line 1218, in _maybe_upcast new_dtype, fill_value = _maybe_promote(dtype, fill_value) File "/.../pandas/core/common.py", line 1124, in _maybe_promote if issubclass(np.dtype(dtype).type, compat.string_types): TypeError: data type not understood

10,770

pandas-dev/pandas

pandas-dev__pandas-10853

d27068f8b78661a64580340a5ab230d0dad17760

testArrayNumpyLabelled fails on Python 2.7.10 ``` FAIL: testArrayNumpyLabelled (pandas.io.tests.test_json.test_ujson.NumpyJSONTests) ---------------------------------------------------------------------- Traceback (most recent call last): File "/Users/ch/repo/pandas/.tox/py27/lib/python2.7/site-packages/pandas/io/tests/test_json/test_ujson.py", line 1141, in testArrayNumpyLabelled self.assertTrue((expectedvals == output[0]).all()) AssertionError: False is not true -------------------- >> begin captured stdout << --------------------- [[42 31] [24 99] [ 2 78]] [[31 42] [99 24] [78 2]] --------------------- >> end captured stdout << ---------------------- ``` Order of dict elements seems to be non-deterministic for python2.7 versions as well.

well this doesn't fail on travis AFAICT, nor have I ever actually seen this fail. So not sure how to repro. Do you have a 32-bit build? This test does seem to depend on a few things, including architecture and the PYTHONHASHSEED environment variable. @kawochen, No, I am running a 64bit interpreter. Yes, the failure is highly hardware/interpreter/seed dependent. But that is the point of this bug report. @cel4 can you give the specific system this fails on. Yes its is somewhat non-deterministic, though in py2 the ordering is deterministic. Kind of a silly test actually (but IIRC came from another package :). Why don't you do a pull-request to sort them first. ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Darwin OS-release: 14.5.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: de_DE.UTF-8 LANG: de_DE.UTF-8 pandas: 0.16.2+340.ge4368de.dirty nose: 1.3.7 Cython: 0.23 numpy: 1.9.2 scipy: None statsmodels: None IPython: None sphinx: None patsy: None dateutil: 2.4.2 pytz: 2015.4 bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: 1.8.6 xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.4.0 html5lib: 0.999999 httplib2: 0.9.1 apiclient: 1.2 sqlalchemy: 1.0.8 pymysql: None psycopg2: None ``` If the test fails or passes depends on the python hashseed. e.g. `PYTHONHASHSEED='1145934480'` deterministically fails that test, `PYTHONHASHSEED='2585294338'` deterministically passes the test here.

2015-08-19T03:59:44Z

[]

Traceback (most recent call last): File "/Users/ch/repo/pandas/.tox/py27/lib/python2.7/site-packages/pandas/io/tests/test_json/test_ujson.py", line 1141, in testArrayNumpyLabelled self.assertTrue((expectedvals == output[0]).all()) AssertionError: False is not true

10,813

pandas-dev/pandas

pandas-dev__pandas-11114

d1feb49267da6074603c6a9dbf6314681378cd81

diff --git a/.travis.yml b/.travis.yml --- a/.travis.yml +++ b/.travis.yml @@ -43,13 +43,6 @@ matrix: - CLIPBOARD_GUI=gtk2 - BUILD_TYPE=conda - DOC_BUILD=true # if rst files were changed, build docs in parallel with tests - - python: 3.3 - env: - - JOB_NAME: "33_nslow" - - NOSE_ARGS="not slow and not disabled" - - FULL_DEPS=true - - CLIPBOARD=xsel - - BUILD_TYPE=conda - python: 3.4 env: - JOB_NAME: "34_nslow" @@ -64,6 +57,13 @@ matrix: - FULL_DEPS=true - CLIPBOARD=xsel - BUILD_TYPE=conda + - python: 3.3 + env: + - JOB_NAME: "33_nslow" + - NOSE_ARGS="not slow and not disabled" + - FULL_DEPS=true + - CLIPBOARD=xsel + - BUILD_TYPE=conda - python: 2.7 env: - JOB_NAME: "27_slow" @@ -104,10 +104,10 @@ matrix: - BUILD_TYPE=pydata - PANDAS_TESTING_MODE="deprecate" allow_failures: - - python: 3.5 + - python: 3.3 env: - - JOB_NAME: "35_nslow" - - NOSE_ARGS="not slow and not network and not disabled" + - JOB_NAME: "33_nslow" + - NOSE_ARGS="not slow and not disabled" - FULL_DEPS=true - CLIPBOARD=xsel - BUILD_TYPE=conda diff --git a/ci/requirements-3.5.txt b/ci/requirements-3.5.txt --- a/ci/requirements-3.5.txt +++ b/ci/requirements-3.5.txt @@ -10,3 +10,15 @@ cython scipy numexpr pytables +html5lib +lxml + +# currently causing some warnings +#sqlalchemy +#pymysql +#psycopg2 + +# not available from conda +#beautiful-soup +#bottleneck +#matplotlib diff --git a/doc/source/install.rst b/doc/source/install.rst --- a/doc/source/install.rst +++ b/doc/source/install.rst @@ -18,7 +18,7 @@ Instructions for installing from source, Python version support ---------------------- -Officially Python 2.6, 2.7, 3.3, and 3.4. +Officially Python 2.6, 2.7, 3.3, 3.4, and 3.5 Installing pandas ----------------- diff --git a/doc/source/release.rst b/doc/source/release.rst --- a/doc/source/release.rst +++ b/doc/source/release.rst @@ -64,6 +64,7 @@ Highlights include: - Support for reading SAS xport files, see :ref:`here <whatsnew_0170.enhancements.sas_xport>` - Documentation comparing SAS to *pandas*, see :ref:`here <compare_with_sas>` - Removal of the automatic TimeSeries broadcasting, deprecated since 0.8.0, see :ref:`here <whatsnew_0170.prior_deprecations>` +- Compatibility with Python 3.5 See the :ref:`v0.17.0 Whatsnew <whatsnew_0170>` overview for an extensive list of all enhancements and bugs that have been fixed in 0.17.0. diff --git a/doc/source/whatsnew/v0.17.0.txt b/doc/source/whatsnew/v0.17.0.txt --- a/doc/source/whatsnew/v0.17.0.txt +++ b/doc/source/whatsnew/v0.17.0.txt @@ -49,6 +49,7 @@ Highlights include: - Support for reading SAS xport files, see :ref:`here <whatsnew_0170.enhancements.sas_xport>` - Documentation comparing SAS to *pandas*, see :ref:`here <compare_with_sas>` - Removal of the automatic TimeSeries broadcasting, deprecated since 0.8.0, see :ref:`here <whatsnew_0170.prior_deprecations>` +- Compatibility with Python 3.5 (:issue:`11097`) Check the :ref:`API Changes <whatsnew_0170.api>` and :ref:`deprecations <whatsnew_0170.deprecations>` before updating. diff --git a/pandas/compat/__init__.py b/pandas/compat/__init__.py --- a/pandas/compat/__init__.py +++ b/pandas/compat/__init__.py @@ -36,9 +36,9 @@ import sys import types -PY3 = (sys.version_info[0] >= 3) PY2 = sys.version_info[0] == 2 - +PY3 = (sys.version_info[0] >= 3) +PY35 = (sys.version_info >= (3, 5)) try: import __builtin__ as builtins diff --git a/pandas/computation/expr.py b/pandas/computation/expr.py --- a/pandas/computation/expr.py +++ b/pandas/computation/expr.py @@ -516,7 +516,54 @@ def visit_Attribute(self, node, **kwargs): raise ValueError("Invalid Attribute context {0}".format(ctx.__name__)) - def visit_Call(self, node, side=None, **kwargs): + def visit_Call_35(self, node, side=None, **kwargs): + """ in 3.5 the starargs attribute was changed to be more flexible, #11097 """ + + if isinstance(node.func, ast.Attribute): + res = self.visit_Attribute(node.func) + elif not isinstance(node.func, ast.Name): + raise TypeError("Only named functions are supported") + else: + try: + res = self.visit(node.func) + except UndefinedVariableError: + # Check if this is a supported function name + try: + res = FuncNode(node.func.id) + except ValueError: + # Raise original error + raise + + if res is None: + raise ValueError("Invalid function call {0}".format(node.func.id)) + if hasattr(res, 'value'): + res = res.value + + if isinstance(res, FuncNode): + + new_args = [ self.visit(arg) for arg in node.args ] + + if node.keywords: + raise TypeError("Function \"{0}\" does not support keyword " + "arguments".format(res.name)) + + return res(*new_args, **kwargs) + + else: + + new_args = [ self.visit(arg).value for arg in node.args ] + + for key in node.keywords: + if not isinstance(key, ast.keyword): + raise ValueError("keyword error in function call " + "'{0}'".format(node.func.id)) + + if key.arg: + kwargs.append(ast.keyword(keyword.arg, self.visit(keyword.value))) + + return self.const_type(res(*new_args, **kwargs), self.env) + + def visit_Call_legacy(self, node, side=None, **kwargs): # this can happen with: datetime.datetime if isinstance(node.func, ast.Attribute): @@ -607,6 +654,13 @@ def visitor(x, y): operands = node.values return reduce(visitor, operands) +# ast.Call signature changed on 3.5, +# conditionally change which methods is named +# visit_Call depending on Python version, #11097 +if compat.PY35: + BaseExprVisitor.visit_Call = BaseExprVisitor.visit_Call_35 +else: + BaseExprVisitor.visit_Call = BaseExprVisitor.visit_Call_legacy _python_not_supported = frozenset(['Dict', 'BoolOp', 'In', 'NotIn']) _numexpr_supported_calls = frozenset(_reductions + _mathops) diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -181,6 +181,7 @@ def build_extensions(self): 'Programming Language :: Python :: 2.7', 'Programming Language :: Python :: 3.3', 'Programming Language :: Python :: 3.4', + 'Programming Language :: Python :: 3.5', 'Programming Language :: Cython', 'Topic :: Scientific/Engineering', ]

TST/COMPAT: python 3.5 compat A few changes are needed: https://travis-ci.org/jreback/pandas/jobs/80331174 (since merged to master) so the Call node as changed need to do something like this: https://bitbucket.org/pytest-dev/pytest/pull-requests/296/astcall-signature-changed-on-35/diff If this is fixed I think everything will pass. I guess this is an API change in python. ``` ====================================================================== ERROR: test_df_use_case (pandas.computation.tests.test_eval.TestMathNumExprPandas) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/travis/build/jreback/pandas/pandas/computation/tests/test_eval.py", line 1488, in test_df_use_case parser=self.parser) File "/home/travis/build/jreback/pandas/pandas/core/frame.py", line 2102, in eval return _eval(expr, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 656, in __init__ self.terms = self.parse() File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 673, in parse return self._visitor.visit(self.expr) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 498, in visit_Assign return self.visit(node.value, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in visit_Call args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in <listcomp> args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 545, in visit_Call if node.starargs is not None: AttributeError: 'Call' object has no attribute 'starargs' ```

2015-09-15T16:32:37Z

[]

Traceback (most recent call last): File "/home/travis/build/jreback/pandas/pandas/computation/tests/test_eval.py", line 1488, in test_df_use_case parser=self.parser) File "/home/travis/build/jreback/pandas/pandas/core/frame.py", line 2102, in eval return _eval(expr, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 656, in __init__ self.terms = self.parse() File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 673, in parse return self._visitor.visit(self.expr) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 498, in visit_Assign return self.visit(node.value, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in visit_Call args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 543, in <listcomp> args = [self.visit(targ) for targ in node.args] File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/travis/build/jreback/pandas/pandas/computation/expr.py", line 545, in visit_Call if node.starargs is not None: AttributeError: 'Call' object has no attribute 'starargs'

10,840

pandas-dev/pandas

pandas-dev__pandas-11366

b09b84e8e0baf89e78b618cdda30af11087d2e4a

diff --git a/doc/source/whatsnew/v0.17.1.txt b/doc/source/whatsnew/v0.17.1.txt --- a/doc/source/whatsnew/v0.17.1.txt +++ b/doc/source/whatsnew/v0.17.1.txt @@ -94,7 +94,7 @@ Bug Fixes - +- Bug in ``pd.eval`` where unary ops in a list error (:issue:`11235`) - Bug in ``squeeze()`` with zero length arrays (:issue:`11230`, :issue:`8999`) diff --git a/pandas/computation/expr.py b/pandas/computation/expr.py --- a/pandas/computation/expr.py +++ b/pandas/computation/expr.py @@ -427,7 +427,7 @@ def visit_Str(self, node, **kwargs): return self.term_type(name, self.env) def visit_List(self, node, **kwargs): - name = self.env.add_tmp([self.visit(e).value for e in node.elts]) + name = self.env.add_tmp([self.visit(e)(self.env) for e in node.elts]) return self.term_type(name, self.env) visit_Tuple = visit_List @@ -655,7 +655,7 @@ def visitor(x, y): return reduce(visitor, operands) # ast.Call signature changed on 3.5, -# conditionally change which methods is named +# conditionally change which methods is named # visit_Call depending on Python version, #11097 if compat.PY35: BaseExprVisitor.visit_Call = BaseExprVisitor.visit_Call_35

Dataframe.eval(): Negative number in list passed to 'in'-expression causes crash on python 3.4.0 The following crashes on python 3.4.0. It works fine on Python 2.7.5. ``` >>> import pandas >>> from io import StringIO >>> data = "foo,bar\n11,12" >>> df = pandas.read_csv(StringIO(data)) >>> df.eval('foo in [11, 32]') 0 True dtype: bool >>> df.eval('foo in [11, -32]') Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/core/frame.py", line 1987, in eval return _eval(expr, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 635, in __init__ self.terms = self.parse() File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 652, in parse return self._visitor.visit(self.expr) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 323, in visit_Expr return self.visit(node.value, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 560, in visit_Compare return self.visit(binop) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 404, in visit_BinOp op, op_class, left, right = self._possibly_transform_eq_ne(node) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 357, in _possibly_transform_eq_ne right = self.visit(node.right, side='right') File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in visit_List name = self.env.add_tmp([self.visit(e).value for e in node.elts]) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in <listcomp> name = self.env.add_tmp([self.visit(e).value for e in node.elts]) AttributeError: 'UnaryOp' object has no attribute 'value' >>> ```

pls show a reproducible example, e.g. show how the actual frame is created so it can be copy-pasted pls show `pd.show_versions()` Sorry, my bad. The lines actually defining the data fell away in during copy-paste. I've updated the example above. Here's show_versions() ``` >>> import pandas >>> pandas.show_versions() INSTALLED VERSIONS ------------------ commit: None python: 3.4.0.final.0 python-bits: 64 OS: Linux OS-release: 3.13.0-48-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 pandas: 0.16.2 nose: None Cython: None numpy: 1.9.3 scipy: None statsmodels: None IPython: None sphinx: None patsy: None dateutil: 2.4.2 pytz: 2015.6 bottleneck: None tables: None numexpr: 2.4.4 matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None ``` yeh I guess this is parsed differently in 3.4 that in 2.7. pull-requests to fix are welcome

2015-10-19T05:40:18Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/core/frame.py", line 1987, in eval return _eval(expr, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/eval.py", line 230, in eval truediv=truediv) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 635, in __init__ self.terms = self.parse() File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 652, in parse return self._visitor.visit(self.expr) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 320, in visit_Module return self.visit(expr, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 323, in visit_Expr return self.visit(node.value, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 560, in visit_Compare return self.visit(binop) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 404, in visit_BinOp op, op_class, left, right = self._possibly_transform_eq_ne(node) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 357, in _possibly_transform_eq_ne right = self.visit(node.right, side='right') File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 314, in visit return visitor(node, **kwargs) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in visit_List name = self.env.add_tmp([self.visit(e).value for e in node.elts]) File "/home/tobias/Envs/qcache-py3/lib/python3.4/site-packages/pandas/computation/expr.py", line 430, in <listcomp> name = self.env.add_tmp([self.visit(e).value for e in node.elts]) AttributeError: 'UnaryOp' object has no attribute 'value'

10,873

pandas-dev/pandas

pandas-dev__pandas-11400

88e8d6e7dfcea3435d7695a0c312690a57c05663

diff --git a/doc/source/whatsnew/v0.17.1.txt b/doc/source/whatsnew/v0.17.1.txt --- a/doc/source/whatsnew/v0.17.1.txt +++ b/doc/source/whatsnew/v0.17.1.txt @@ -70,7 +70,7 @@ Bug Fixes - Bug in ``HDFStore.append`` with strings whose encoded length exceded the max unencoded length (:issue:`11234`) - Bug in merging ``datetime64[ns, tz]`` dtypes (:issue:`11405`) - Bug in ``HDFStore.select`` when comparing with a numpy scalar in a where clause (:issue:`11283`) - +- Bug in using ``DataFrame.ix`` with a multi-index indexer(:issue:`11372`) - Bug in tz-conversions with an ambiguous time and ``.dt`` accessors (:issue:`11295`) diff --git a/pandas/core/indexing.py b/pandas/core/indexing.py --- a/pandas/core/indexing.py +++ b/pandas/core/indexing.py @@ -443,11 +443,14 @@ def can_do_equal_len(): # we have an equal len Frame if isinstance(value, ABCDataFrame) and value.ndim > 1: sub_indexer = list(indexer) + multiindex_indexer = isinstance(labels, MultiIndex) for item in labels: if item in value: sub_indexer[info_axis] = item - v = self._align_series(tuple(sub_indexer), value[item]) + v = self._align_series( + tuple(sub_indexer), value[item], multiindex_indexer + ) else: v = np.nan @@ -516,8 +519,28 @@ def can_do_equal_len(): self.obj._data = self.obj._data.setitem(indexer=indexer, value=value) self.obj._maybe_update_cacher(clear=True) - def _align_series(self, indexer, ser): - # indexer to assign Series can be tuple, slice, scalar + def _align_series(self, indexer, ser, multiindex_indexer=False): + """ + Parameters + ---------- + indexer : tuple, slice, scalar + The indexer used to get the locations that will be set to + `ser` + + ser : pd.Series + The values to assign to the locations specified by `indexer` + + multiindex_indexer : boolean, optional + Defaults to False. Should be set to True if `indexer` was from + a `pd.MultiIndex`, to avoid unnecessary broadcasting. + + + Returns: + -------- + `np.array` of `ser` broadcast to the appropriate shape for assignment + to the locations selected by `indexer` + + """ if isinstance(indexer, (slice, np.ndarray, list, Index)): indexer = tuple([indexer]) @@ -555,7 +578,7 @@ def _align_series(self, indexer, ser): ser = ser.reindex(obj.axes[0][indexer[0]], copy=True)._values # single indexer - if len(indexer) > 1: + if len(indexer) > 1 and not multiindex_indexer: l = len(indexer[1]) ser = np.tile(ser, l).reshape(l, -1).T

DataFrame.ix[idx, :] = value sets wrong values when idx is a MultiIndex and DataFrame.columns is also a MultiIndex This code is broken in `0.17.0` but not in `0.15.2`: ``` python import pandas as pd import numpy as np np.random.seed(1) from itertools import product from pandas.util.testing import assert_frame_equal pd.show_versions() idx = pd.MultiIndex.from_tuples( list( product(['A', 'B', 'C'], pd.date_range('2015-01-01', '2015-04-01', freq='MS')) ) ) sub = pd.MultiIndex.from_tuples( [('A', pd.Timestamp('2015-01-01')), ('A', pd.Timestamp('2015-02-01'))] ) # if cols = ['foo', 'bar', 'baz', 'quux'], there is no error. cols = pd.MultiIndex.from_tuples( list( product(['foo', 'bar'], pd.date_range('2015-01-01', '2015-02-01', freq='MS')) ) ) test = pd.DataFrame(np.random.random((12, 4)), index=idx, columns=cols) vals = pd.DataFrame(np.random.random((2, 4)), index=sub, columns=cols) test.ix[sub, :] = vals print test.ix[sub, :] print vals assert_frame_equal(test.ix[sub, :], vals) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.287775 0.130029 0.019367 0.678836 foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.211628 0.265547 0.491573 0.053363 Traceback (most recent call last): File "c:\dev\code\sandbox\multiindex.py", line 41, in <module> assert_frame_equal(test.ix[sub, :], vals) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 1028, in assert_frame_equal obj='DataFrame.iloc[:, {0}]'.format(i)) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 925, in assert_series_equal check_less_precise, obj='{0}'.format(obj)) File "pandas\src\testing.pyx", line 58, in pandas._testing.assert_almost_equal (pandas\src\testing.c:3809) File "pandas\src\testing.pyx", line 147, in pandas._testing.assert_almost_equal (pandas\src\testing.c:2685) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 798, in raise_assert_detail raise AssertionError(msg) AssertionError: DataFrame.iloc[:, 0] are different DataFrame.iloc[:, 0] values are different (50.0 %) [left]: [0.287775338586, 0.287775338586] [right]: [0.287775338586, 0.211628116] ``` ### 0.15.2 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: en_GB pandas: 0.15.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: 0.15.1 statsmodels: None IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.3.0 dateutil: 2.4.1 pytz: 2015.4 bottleneck: 1.0.0 tables: 3.2.0 numexpr: 2.4.3 matplotlib: 1.4.3 openpyxl: 1.8.5 xlrd: 0.9.4 xlwt: 0.7.5 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None rpy2: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.211628 0.265547 0.491573 0.053363 foo bar 2015-01-01 2015-02-01 2015-01-01 2015-02-01 A 2015-01-01 0.287775 0.130029 0.019367 0.678836 2015-02-01 0.211628 0.265547 0.491573 0.053363 ```

Indexing with a specific set of columns also gives the error: Code sample: ``` python import pandas as pd import numpy as np np.random.seed(1) from itertools import product from pandas.util.testing import assert_frame_equal pd.show_versions() idx = pd.MultiIndex.from_tuples( list( product(['A', 'B', 'C'], pd.date_range('2015-01-01', '2015-04-01', freq='MS')) ) ) cols = pd.MultiIndex.from_tuples( list( product(['foo', 'bar'], pd.date_range('2016-01-01', '2016-02-01', freq='MS')) ) ) # if cols = ['foo', 'bar', 'baz', 'quux'], there is no error. test = pd.DataFrame(np.random.random((12, 4)), index=idx, columns=cols) subidx = pd.MultiIndex.from_tuples( [('A', pd.Timestamp('2015-01-01')), ('A', pd.Timestamp('2015-02-01'))] ) subcols = pd.MultiIndex.from_tuples( [('foo', pd.Timestamp('2016-01-01')), ('foo', pd.Timestamp('2016-02-01'))] ) vals = pd.DataFrame(np.random.random((2, 2)), index=subidx, columns=subcols) test.ix[subidx, subcols] = vals print test.ix[subidx, subcols] print vals assert_frame_equal(test.ix[subidx, subcols], vals) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None foo 2016-01-01 2016-02-01 A 2015-01-01 0.287775 0.130029 2015-02-01 0.287775 0.130029 foo 2016-01-01 2016-02-01 A 2015-01-01 0.287775 0.130029 2015-02-01 0.019367 0.678836 Traceback (most recent call last): File "c:\dev\code\sandbox\multiindex.py", line 48, in <module> assert_frame_equal(test.ix[subidx, subcols], vals) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 1028, in assert_frame_equal obj='DataFrame.iloc[:, {0}]'.format(i)) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 925, in assert_series_equal check_less_precise, obj='{0}'.format(obj)) File "pandas\src\testing.pyx", line 58, in pandas._testing.assert_almost_equal (pandas\src\testing.c:3809) File "pandas\src\testing.pyx", line 147, in pandas._testing.assert_almost_equal (pandas\src\testing.c:2685) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 798, in raise_assert_detail raise AssertionError(msg) AssertionError: DataFrame.iloc[:, 0] are different DataFrame.iloc[:, 0] values are different (50.0 %) [left]: [0.287775338586, 0.287775338586] [right]: [0.287775338586, 0.0193669578703] ``` (Deleted- misread something, my previous suggestion was not really a fix) hmm, surprised that broke. there is not must testing on that sub-section actually The issue is here: https://github.com/pydata/pandas/blob/master/pandas/core/indexing.py#L450 `self._align_series` is called on a sub-section of the frame, but in the aligner, it looks at it and says oh you are a frame so gives back the wrong result. So could prob pass in an additional parameter which would determine this. Since I've already got two test cases, I'd be happy to have a go if I can be pointed in the right direction. I'll start by looking at the history of `indexing.py` and following any referenced issues / PRs the pointer above is to the relevant issues. the way to do this is to setup the test cases and the expected results (in test_indexing); they should fail before a fix, then you can step thru to see where to put a fix and go from there

2015-10-21T08:40:16Z

[]

Traceback (most recent call last): File "c:\dev\code\sandbox\multiindex.py", line 41, in <module> assert_frame_equal(test.ix[sub, :], vals) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 1028, in assert_frame_equal obj='DataFrame.iloc[:, {0}]'.format(i)) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 925, in assert_series_equal check_less_precise, obj='{0}'.format(obj)) File "pandas\src\testing.pyx", line 58, in pandas._testing.assert_almost_equal (pandas\src\testing.c:3809) File "pandas\src\testing.pyx", line 147, in pandas._testing.assert_almost_equal (pandas\src\testing.c:2685) File "c:\python\envs\pd017\lib\site-packages\pandas\util\testing.py", line 798, in raise_assert_detail raise AssertionError(msg) AssertionError: DataFrame.iloc[:, 0] are different

10,878

pandas-dev/pandas

pandas-dev__pandas-11427

faa6cc744ba6086ddcef66c462823a169e1a733c

diff --git a/doc/source/timeseries.rst b/doc/source/timeseries.rst --- a/doc/source/timeseries.rst +++ b/doc/source/timeseries.rst @@ -954,6 +954,52 @@ These can be used as arguments to ``date_range``, ``bdate_range``, constructors for ``DatetimeIndex``, as well as various other timeseries-related functions in pandas. +Anchored Offset Semantics +~~~~~~~~~~~~~~~~~~~~~~~~~ + +For those offsets that are anchored to the start or end of specific +frequency (``MonthEnd``, ``MonthBegin``, ``WeekEnd``, etc) the following +rules apply to rolling forward and backwards. + +When ``n`` is not 0, if the given date is not on an anchor point, it snapped to the next(previous) +anchor point, and moved ``|n|-1`` additional steps forwards or backwards. + +.. ipython:: python + + pd.Timestamp('2014-01-02') + MonthBegin(n=1) + pd.Timestamp('2014-01-02') + MonthEnd(n=1) + + pd.Timestamp('2014-01-02') - MonthBegin(n=1) + pd.Timestamp('2014-01-02') - MonthEnd(n=1) + + pd.Timestamp('2014-01-02') + MonthBegin(n=4) + pd.Timestamp('2014-01-02') - MonthBegin(n=4) + +If the given date *is* on an anchor point, it is moved ``|n|`` points forwards +or backwards. + +.. ipython:: python + + pd.Timestamp('2014-01-01') + MonthBegin(n=1) + pd.Timestamp('2014-01-31') + MonthEnd(n=1) + + pd.Timestamp('2014-01-01') - MonthBegin(n=1) + pd.Timestamp('2014-01-31') - MonthEnd(n=1) + + pd.Timestamp('2014-01-01') + MonthBegin(n=4) + pd.Timestamp('2014-01-31') - MonthBegin(n=4) + +For the case when ``n=0``, the date is not moved if on an anchor point, otherwise +it is rolled forward to the next anchor point. + +.. ipython:: python + + pd.Timestamp('2014-01-02') + MonthBegin(n=0) + pd.Timestamp('2014-01-02') + MonthEnd(n=0) + + pd.Timestamp('2014-01-01') + MonthBegin(n=0) + pd.Timestamp('2014-01-31') + MonthEnd(n=0) + .. _timeseries.legacyaliases: Legacy Aliases diff --git a/doc/source/whatsnew/v0.18.0.txt b/doc/source/whatsnew/v0.18.0.txt --- a/doc/source/whatsnew/v0.18.0.txt +++ b/doc/source/whatsnew/v0.18.0.txt @@ -190,7 +190,7 @@ Bug Fixes - + - Bug in vectorized ``DateOffset`` when ``n`` parameter is ``0`` (:issue:`11370`) diff --git a/pandas/tseries/offsets.py b/pandas/tseries/offsets.py --- a/pandas/tseries/offsets.py +++ b/pandas/tseries/offsets.py @@ -444,7 +444,7 @@ def _beg_apply_index(self, i, freq): from pandas.tseries.frequencies import get_freq_code base, mult = get_freq_code(freq) base_period = i.to_period(base) - if self.n < 0: + if self.n <= 0: # when subtracting, dates on start roll to prior roll = np.where(base_period.to_timestamp() == i - off, self.n, self.n + 1) @@ -464,7 +464,7 @@ def _end_apply_index(self, i, freq): base, mult = get_freq_code(freq) base_period = i.to_period(base) if self.n > 0: - # when adding, dtates on end roll to next + # when adding, dates on end roll to next roll = np.where(base_period.to_timestamp(how='end') == i - off, self.n, self.n - 1) else: @@ -1081,8 +1081,7 @@ def apply(self, other): @apply_index_wraps def apply_index(self, i): - months = self.n - 1 if self.n >= 0 else self.n - shifted = tslib.shift_months(i.asi8, months, 'end') + shifted = tslib.shift_months(i.asi8, self.n, 'end') return i._shallow_copy(shifted) def onOffset(self, dt): @@ -1108,8 +1107,7 @@ def apply(self, other): @apply_index_wraps def apply_index(self, i): - months = self.n + 1 if self.n < 0 else self.n - shifted = tslib.shift_months(i.asi8, months, 'start') + shifted = tslib.shift_months(i.asi8, self.n, 'start') return i._shallow_copy(shifted) def onOffset(self, dt): @@ -1777,6 +1775,7 @@ def apply(self, other): @apply_index_wraps def apply_index(self, i): freq_month = 12 if self.startingMonth == 1 else self.startingMonth - 1 + # freq_month = self.startingMonth freqstr = 'Q-%s' % (_int_to_month[freq_month],) return self._beg_apply_index(i, freqstr) diff --git a/pandas/tslib.pyx b/pandas/tslib.pyx --- a/pandas/tslib.pyx +++ b/pandas/tslib.pyx @@ -4458,7 +4458,8 @@ def shift_months(int64_t[:] dtindex, int months, object day=None): Py_ssize_t i pandas_datetimestruct dts int count = len(dtindex) - int days_in_current_month + int months_to_roll + bint roll_check int64_t[:] out = np.empty(count, dtype='int64') if day is None: @@ -4472,36 +4473,44 @@ def shift_months(int64_t[:] dtindex, int months, object day=None): dts.day = min(dts.day, days_in_month(dts)) out[i] = pandas_datetimestruct_to_datetime(PANDAS_FR_ns, &dts) elif day == 'start': + roll_check = False + if months <= 0: + months += 1 + roll_check = True with nogil: for i in range(count): if dtindex[i] == NPY_NAT: out[i] = NPY_NAT; continue pandas_datetime_to_datetimestruct(dtindex[i], PANDAS_FR_ns, &dts) - dts.year = _year_add_months(dts, months) - dts.month = _month_add_months(dts, months) + months_to_roll = months + + # offset semantics - if on the anchor point and going backwards + # shift to next + if roll_check and dts.day == 1: + months_to_roll -= 1 + + dts.year = _year_add_months(dts, months_to_roll) + dts.month = _month_add_months(dts, months_to_roll) + dts.day = 1 - # offset semantics - when subtracting if at the start anchor - # point, shift back by one more month - if months <= 0 and dts.day == 1: - dts.year = _year_add_months(dts, -1) - dts.month = _month_add_months(dts, -1) - else: - dts.day = 1 out[i] = pandas_datetimestruct_to_datetime(PANDAS_FR_ns, &dts) elif day == 'end': + roll_check = False + if months > 0: + months -= 1 + roll_check = True with nogil: for i in range(count): if dtindex[i] == NPY_NAT: out[i] = NPY_NAT; continue pandas_datetime_to_datetimestruct(dtindex[i], PANDAS_FR_ns, &dts) - days_in_current_month = days_in_month(dts) - - dts.year = _year_add_months(dts, months) - dts.month = _month_add_months(dts, months) + months_to_roll = months # similar semantics - when adding shift forward by one # month if already at an end of month - if months >= 0 and dts.day == days_in_current_month: - dts.year = _year_add_months(dts, 1) - dts.month = _month_add_months(dts, 1) + if roll_check and dts.day == days_in_month(dts): + months_to_roll += 1 + + dts.year = _year_add_months(dts, months_to_roll) + dts.month = _month_add_months(dts, months_to_roll) dts.day = days_in_month(dts) out[i] = pandas_datetimestruct_to_datetime(PANDAS_FR_ns, &dts)

Vectorised addition of MonthOffset(n=0) returns different values to item-by-item addition This code returns different values in `0.17.0` and `0.15.2` ``` python import pandas as pd from pandas.util.testing import assert_index_equal pd.show_versions() offsets = [ pd.offsets.Day, pd.offsets.MonthBegin, pd.offsets.QuarterBegin, pd.offsets.YearBegin, ] dates = pd.date_range('2011-01-01', '2011-01-05', freq='D') for offset in offsets: # adding each item individually or vectorised should give same answer expected_vec = dates + offset(n=0) expected = pd.DatetimeIndex([d + offset(n=0) for d in dates]) msg = "offset: {}, vectorised: {}, individual: {}".format( offset, expected_vec, expected ) try: if pd.__version__ == '0.17.0': assert_index_equal(expected_vec, expected, check_names=False) else: assert_index_equal(expected_vec, expected) except AssertionError as er: raise Exception(msg + str(er)) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None Traceback (most recent call last): File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 24, in <module> raise Exception(msg + str(er)) Exception: offset: <class 'pandas.tseries.offsets.MonthBegin'>, vectorised: DatetimeIndex(['2010-12-01', '2011-01-01', '2011-01-01', '2011-01-01', '2011-01-01'], dtype='datetime64[ns]', freq=None), individual: DatetimeIndex(['2011-01-01', '2011-02-01', '2011-02-01', '2011-02-01', '2011-02-01'], dtype='datetime64[ns]', freq=None)Index are different Index values are different (100.0 %) [left]: DatetimeIndex(['2010-12-01', '2011-01-01', '2011-01-01', '2011-01-01', '2011-01-01'], dtype='datetime64[ns]', freq=None) [right]: DatetimeIndex(['2011-01-01', '2011-02-01', '2011-02-01', '2011-02-01', '2011-02-01'], dtype='datetime64[ns]', freq=None) ``` ### 0.15.2 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: en_GB pandas: 0.15.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: 0.15.1 statsmodels: None IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.3.0 dateutil: 2.4.1 pytz: 2015.4 bottleneck: 1.0.0 tables: 3.2.0 numexpr: 2.4.3 matplotlib: 1.4.3 openpyxl: 1.8.5 xlrd: 0.9.4 xlwt: 0.7.5 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None rpy2: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None ```

This is from #10744, I didn't have the n=0 semantics right (and apparently didn't test!). It'll be a couple days, but I'll submit a fix. Many thanks for quick response! `MonthEnd` also not working: Test script ``` python import pandas as pd from pandas.util.testing import assert_index_equal pd.show_versions() offsets = [ pd.offsets.MonthEnd, pd.offsets.QuarterEnd, pd.offsets.YearEnd, ] dates = pd.date_range('2011-01-01', '2011-01-05', freq='D') for offset in offsets: # adding each item individually or vectorised should give same answer expected_vec = dates + offset(n=0) expected = pd.DatetimeIndex([d + offset(n=0) for d in dates]) msg = "offset: {}, vectorised: {}, individual: {}".format( offset, expected_vec, expected ) try: if pd.__version__ == '0.17.0': assert_index_equal(expected_vec, expected, check_names=False) else: assert_index_equal(expected_vec, expected) except AssertionError as er: raise Exception(msg + str(er)) ``` ### 0.17.0 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.17.0 nose: 1.3.7 pip: 7.1.0 setuptools: 18.0.1 Cython: 0.22 numpy: 1.10.1 scipy: 0.16.0 statsmodels: 0.6.1 IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.4.0 dateutil: 2.4.1 pytz: 2015.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.4.4 matplotlib: 1.4.3 openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: 0.7.3 lxml: None bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None Traceback (most recent call last): File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 34, in <module> raise Exception(msg + str(er)) Exception: offset: <class 'pandas.tseries.offsets.MonthEnd'>, vectorised: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31'], dtype='datetime64[ns]', freq=None), individual: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31'], dtype='datetime64[ns]', freq=None)Index are different Index values are different (100.0 %) [left]: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31'], dtype='datetime64[ns]', freq=None) [right]: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31'], dtype='datetime64[ns]', freq=None) ``` ### 0.15.2 ``` python INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel byteorder: little LC_ALL: None LANG: en_GB pandas: 0.15.2 nose: 1.3.7 Cython: 0.22 numpy: 1.9.2 scipy: 0.15.1 statsmodels: None IPython: 3.2.1 sphinx: 1.3.1 patsy: 0.3.0 dateutil: 2.4.1 pytz: 2015.4 bottleneck: 1.0.0 tables: 3.2.0 numexpr: 2.4.3 matplotlib: 1.4.3 openpyxl: 1.8.5 xlrd: 0.9.4 xlwt: 0.7.5 xlsxwriter: 0.7.3 lxml: 3.4.4 bs4: 4.3.2 html5lib: 0.999 httplib2: None apiclient: None rpy2: None sqlalchemy: 1.0.7 pymysql: None psycopg2: None ``` Probably also wrong for `YearEnd` and `QuarterEnd` too as the counting logic is shared IIRC. On Mon, Oct 19, 2015 at 10:31 AM, Petra Chong notifications@github.com wrote: > MonthEnd also not working: > > Test script > > import pandas as pdfrom pandas.util.testing import assert_index_equal > > pd.show_versions() > > offsets = [ > pd.offsets.MonthEnd, > pd.offsets.QuarterEnd, pd.offsets.YearEnd, > ] > > dates = pd.date_range('2011-01-01', '2011-01-05', freq='D') > for offset in offsets: > # adding each item individually or vectorised should give same answer > expected_vec = dates + offset(n=0) > expected = pd.DatetimeIndex([d + offset(n=0) for d in dates]) > > ``` > msg = "offset: {}, vectorised: {}, individual: {}".format( > offset, expected_vec, expected > ) > try: > if pd.__version__ == '0.17.0': > assert_index_equal(expected_vec, expected, check_names=False) > else: > assert_index_equal(expected_vec, expected) > except AssertionError as er: > raise Exception(msg + str(er)) > ``` > > 0.17.0 > > INSTALLED VERSIONS------------------ > commit: None > python: 2.7.10.final.0 > python-bits: 64 > OS: Windows > OS-release: 7 > machine: AMD64 > processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel > byteorder: little > LC_ALL: None > LANG: None > > pandas: 0.17.0 > nose: 1.3.7 > pip: 7.1.0 > setuptools: 18.0.1 > Cython: 0.22 > numpy: 1.10.1 > scipy: 0.16.0 > statsmodels: 0.6.1 > IPython: 3.2.1 > sphinx: 1.3.1 > patsy: 0.4.0 > dateutil: 2.4.1 > pytz: 2015.4 > blosc: None > bottleneck: 1.0.0 > tables: 3.2.2 > numexpr: 2.4.4 > matplotlib: 1.4.3 > openpyxl: None > xlrd: 0.9.4 > xlwt: None > xlsxwriter: 0.7.3 > lxml: None > bs4: 4.3.2 > html5lib: 0.999 > httplib2: None > apiclient: None > sqlalchemy: 1.0.7 > pymysql: None > psycopg2: None > Traceback (most recent call last): > File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 34, in <module> > raise Exception(msg + str(er))Exception: offset: <class 'pandas.tseries.offsets.MonthEnd'>, vectorised: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', > '2010-12-31'], > dtype='datetime64[ns]', freq=None), individual: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', > '2011-01-31'], > dtype='datetime64[ns]', freq=None)Index are different > > Index values are different (100.0 %) > [left]: DatetimeIndex(['2010-12-31', '2010-12-31', '2010-12-31', '2010-12-31', > '2010-12-31'], > dtype='datetime64[ns]', freq=None) > [right]: DatetimeIndex(['2011-01-31', '2011-01-31', '2011-01-31', '2011-01-31', > '2011-01-31'], > dtype='datetime64[ns]', freq=None) > > 0.15.2 > > INSTALLED VERSIONS------------------ > commit: None > python: 2.7.10.final.0 > python-bits: 64 > OS: Windows > OS-release: 7 > machine: AMD64 > processor: Intel64 Family 6 Model 26 Stepping 5, GenuineIntel > byteorder: little > LC_ALL: None > LANG: en_GB > > pandas: 0.15.2 > nose: 1.3.7 > Cython: 0.22 > numpy: 1.9.2 > scipy: 0.15.1 > statsmodels: None > IPython: 3.2.1 > sphinx: 1.3.1 > patsy: 0.3.0 > dateutil: 2.4.1 > pytz: 2015.4 > bottleneck: 1.0.0 > tables: 3.2.0 > numexpr: 2.4.3 > matplotlib: 1.4.3 > openpyxl: 1.8.5 > xlrd: 0.9.4 > xlwt: 0.7.5 > xlsxwriter: 0.7.3 > lxml: 3.4.4 > bs4: 4.3.2 > html5lib: 0.999 > httplib2: None > apiclient: None > rpy2: None > sqlalchemy: 1.0.7 > pymysql: None > psycopg2: None > > — > Reply to this email directly or view it on GitHub > https://github.com/pydata/pandas/issues/11370#issuecomment-149250402.

2015-10-24T21:09:45Z

[]

Traceback (most recent call last): File "c:\dev\code\sandbox\pandas_17_vs_15_dateoffsets.py", line 24, in <module> raise Exception(msg + str(er)) Exception: offset: <class 'pandas.tseries.offsets.MonthBegin'>, vectorised: DatetimeIndex(['2010-12-01', '2011-01-01', '2011-01-01', '2011-01-01',

10,881

pandas-dev/pandas

pandas-dev__pandas-11653

2d038327b7ea805a2d9c3db9ca3dd2b459e694bb

diff --git a/doc/source/whatsnew/v0.17.1.txt b/doc/source/whatsnew/v0.17.1.txt --- a/doc/source/whatsnew/v0.17.1.txt +++ b/doc/source/whatsnew/v0.17.1.txt @@ -169,7 +169,7 @@ Bug Fixes - +- Bug in indexing with a ``range``, (:issue:`11652`) - Bug in ``to_sql`` using unicode column names giving UnicodeEncodeError with (:issue:`11431`). diff --git a/pandas/core/index.py b/pandas/core/index.py --- a/pandas/core/index.py +++ b/pandas/core/index.py @@ -1755,7 +1755,8 @@ def get_loc(self, key, method=None, tolerance=None): if tolerance is not None: raise ValueError('tolerance argument only valid if using pad, ' 'backfill or nearest lookups') - return self._engine.get_loc(_values_from_object(key)) + key = _values_from_object(key) + return self._engine.get_loc(key) indexer = self.get_indexer([key], method=method, tolerance=tolerance) diff --git a/pandas/core/indexing.py b/pandas/core/indexing.py --- a/pandas/core/indexing.py +++ b/pandas/core/indexing.py @@ -104,6 +104,8 @@ def _get_setitem_indexer(self, key): if isinstance(key, tuple) and not self.ndim < len(key): return self._convert_tuple(key, is_setter=True) + if isinstance(key, range): + return self._convert_range(key, is_setter=True) try: return self._convert_to_indexer(key, is_setter=True) @@ -156,6 +158,10 @@ def _convert_tuple(self, key, is_setter=False): keyidx.append(idx) return tuple(keyidx) + def _convert_range(self, key, is_setter=False): + """ convert a range argument """ + return list(key) + def _convert_scalar_indexer(self, key, axis): # if we are accessing via lowered dim, use the last dim ax = self.obj._get_axis(min(axis, self.ndim - 1))

BUG: ValueError when indexing using range only when length >= 1,000,000 From SO (http://stackoverflow.com/questions/33814223/strange-error-in-pandas-indexing-with-range-when-length-1-000-000) Pandas raises a ValueError when assigning multiple values to a Series (or DataFrame) using range(x) where x > 1. This error is raised only when its length is one million or larger. ``` python import pandas as pd for x in [5, 999999, 1000000]: s = pd.Series(index=range(x)) print('series length =', len(s)) # assigning value with range(1), always works s.loc[range(1)] = 42 # reading values with range(x>1), always works _ = s.loc[range(2)] # assigning values with range(x>1), fails only when len >= 1 million s.loc[range(2)] = 42 ``` Output: ``` python series length = 5 series length = 999999 series length = 1000000 Traceback (most recent call last): File "<stdin>", line 9, in <module> File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 114, in __setitem__ indexer = self._get_setitem_indexer(key) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 109, in _get_setitem_indexer return self._convert_to_indexer(key, is_setter=True) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 1042, in _convert_to_indexer return labels.get_loc(obj) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/index.py", line 1692, in get_loc return self._engine.get_loc(_values_from_object(key)) File "pandas/index.pyx", line 137, in pandas.index.IndexEngine.get_loc (pandas/index.c:3979) File "pandas/index.pyx", line 145, in pandas.index.IndexEngine.get_loc (pandas/index.c:3680) File "pandas/index.pyx", line 464, in pandas.index._bin_search (pandas/index.c:9124) ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() ``` Tested on pandas 0.17.0 and python 3.4.

odd, only happens on py3. `range(2)` is a list in py2.x. `s.loc[list[range(2)]] = 42` works fine with py3, too. It seems to be failing only with `range` object.

2015-11-19T22:47:48Z

[]

Traceback (most recent call last): File "<stdin>", line 9, in <module> File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 114, in __setitem__ indexer = self._get_setitem_indexer(key) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 109, in _get_setitem_indexer return self._convert_to_indexer(key, is_setter=True) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/indexing.py", line 1042, in _convert_to_indexer return labels.get_loc(obj) File "/home/nekobon/.env_exp/lib/python3.4/site-packages/pandas/core/index.py", line 1692, in get_loc return self._engine.get_loc(_values_from_object(key)) File "pandas/index.pyx", line 137, in pandas.index.IndexEngine.get_loc (pandas/index.c:3979) File "pandas/index.pyx", line 145, in pandas.index.IndexEngine.get_loc (pandas/index.c:3680) File "pandas/index.pyx", line 464, in pandas.index._bin_search (pandas/index.c:9124) ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

10,898

pandas-dev/pandas

pandas-dev__pandas-11714

547750aa5ba5b4a1b5d0cde05cc21e588b30cc27

diff --git a/doc/source/whatsnew/v0.18.0.txt b/doc/source/whatsnew/v0.18.0.txt --- a/doc/source/whatsnew/v0.18.0.txt +++ b/doc/source/whatsnew/v0.18.0.txt @@ -32,6 +32,7 @@ Other enhancements ^^^^^^^^^^^^^^^^^^ - Handle truncated floats in SAS xport files (:issue:`11713`) +- ``read_excel`` now supports s3 urls of the format ``s3://bucketname/filename`` (:issue:`11447`) .. _whatsnew_0180.enhancements.rounding: diff --git a/pandas/io/excel.py b/pandas/io/excel.py --- a/pandas/io/excel.py +++ b/pandas/io/excel.py @@ -11,7 +11,7 @@ from pandas.core.frame import DataFrame from pandas.io.parsers import TextParser -from pandas.io.common import _is_url, _urlopen, _validate_header_arg +from pandas.io.common import _is_url, _urlopen, _validate_header_arg, get_filepath_or_buffer, _is_s3_url from pandas.tseries.period import Period from pandas import json from pandas.compat import (map, zip, reduce, range, lrange, u, add_metaclass, @@ -199,7 +199,10 @@ def __init__(self, io, **kwds): raise ValueError("Unknown engine: %s" % engine) if isinstance(io, compat.string_types): - if _is_url(io): + if _is_s3_url(io): + buffer, _, _ = get_filepath_or_buffer(io) + self.book = xlrd.open_workbook(file_contents=buffer.read()) + elif _is_url(io): data = _urlopen(io).read() self.book = xlrd.open_workbook(file_contents=data) else:

ENH read_excel error when accessing AWS S3 URL Summary: read_excel is unable to read a file using the same S3 URL syntax as read_csv. read_excel should support accessing S3 data in the same manner as read_csv read_excel fails with the following error: ``` python >>> import pandas as pd >>> df = pd.read_excel("s3://my-bucket/my_file.xlsx") Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 163, in read_excel io = ExcelFile(io, engine=engine) File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 206, in __init__ self.book = xlrd.open_workbook(io) File "/usr/local/lib/python2.6/site-packages/xlrd/__init__.py", line 394, in open_workbook f = open(filename, "rb") IOError: [Errno 2] No such file or directory: 's3://my-bucket/my_file.xlsx' >>> ``` read_csv on the other hand is able to successfully read a csv file in the same S3 bucket using the same URL syntax: ``` python >>> import pandas as pd >>> df = pd.read_csv("s3://my-bucket/my_file.csv") >>> len(df.index) 1187 >>> ``` For the record, read_csv can also see the xlsx file but returns parse errors when attempting to tokenize the data. ``` python >>> import pandas as pd >>> df = pd.read_csv("s3://my-bucket/my_file.xlsx") Exception pandas.parser.CParserError: CParserError('Error tokenizing data. C error: Expected 9 fields in line 210, saw 10\n',) in 'pandas.parser.TextReader._tokenize_rows' ignored >>> ``` read_excel successfully reads and parses a local copy of the xlsx file ``` python >>> import pandas as pd >>> df = pd.read_excel("my_file.xlsx") >>> len(df.index) 221 >>> ``` Pandas version string and dependencies: ``` python >>> pd.show_versions() INSTALLED VERSIONS ------------------ commit: None python: 2.6.9.final.0 python-bits: 64 OS: Linux OS-release: 3.14.48-33.39.amzn1.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 pandas: 0.17.0 nose: 1.3.4 pip: 6.1.1 setuptools: 12.2 Cython: None numpy: 1.10.1 scipy: 0.16.0 statsmodels: None IPython: None sphinx: None patsy: None dateutil: 2.4.2 pytz: 2015.7 blosc: None bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: None xlrd: 0.9.4 xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None >>> ```

This is almost a trivial enhancement, just add `_is_s3_url` [here](https://github.com/pydata/pandas/blob/master/pandas/io/excel.py#L202) post a file link as an example and i'll put it on our test s3 bucket.

2015-11-27T20:09:28Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 163, in read_excel io = ExcelFile(io, engine=engine) File "/usr/local/lib64/python2.6/site-packages/pandas/io/excel.py", line 206, in __init__ self.book = xlrd.open_workbook(io) File "/usr/local/lib/python2.6/site-packages/xlrd/__init__.py", line 394, in open_workbook f = open(filename, "rb") IOError: [Errno 2] No such file or directory: 's3://my-bucket/my_file.xlsx'

10,901

pandas-dev/pandas

pandas-dev__pandas-12043

1ae6384a0c04be8a1faddaa85751a9cac2f5a42a

diff --git a/pandas/tseries/index.py b/pandas/tseries/index.py --- a/pandas/tseries/index.py +++ b/pandas/tseries/index.py @@ -1804,7 +1804,6 @@ def indexer_between_time(self, start_time, end_time, include_start=True, "%I%M%S%p") include_start : boolean, default True include_end : boolean, default True - tz : string or pytz.timezone or dateutil.tz.tzfile, default None Returns -------

Error in doc of DatetimeIndex.indexer_between_time Hello everyone, This is not a major issue: I was trying to use the `tz` parameter as indicated on the [documentation](http://pandas.pydata.org/pandas-docs/version/0.17.1/generated/pandas.DatetimeIndex.indexer_between_time.html?highlight=indexer_between_time#pandas.DatetimeIndex.indexer_between_time) of `DatetimeIndex.indexer_between_time` and it turns out that `tz` is not implemented. Here is an example: ``` python import pandas as pd pd.DatetimeIndex(['2016-01-01 00:00:00', '2016-01-01 01:00:00', '2016-01-01 02:00:00']).indexer_between_time('01:00', '02:00', tz='Europe/Paris') ``` ``` Traceback (most recent call last): File "~/anaconda/lib/python2.7/site-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-98-258497b42264>", line 1, in <module> pd.DatetimeIndex(['2016-01-01 00:00:00', '2016-01-01 01:00:00', '2016-01-01 02:00:00']).indexer_between_time('01:00', '02:00', tz='Europe/Paris') TypeError: indexer_between_time() got an unexpected keyword argument 'tz' ```

That doc should be fixed (and you see the signature no longer has a tz argument). Having a `tz` makes no sense in this context. @joseRLC want to do a pull-request for that? Can I try this out? This seems like a good issue for a first PR. @RahulHP Certainly, go for it! If you have any questions about the workflow (see http://pandas.pydata.org/pandas-docs/stable/contributing.html), just ask (you can also use the gitter channel for that: https://gitter.im/pydata/pandas)

2016-01-15T08:44:51Z

[]

Traceback (most recent call last): File "~/anaconda/lib/python2.7/site-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-98-258497b42264>", line 1, in <module> pd.DatetimeIndex(['2016-01-01 00:00:00', '2016-01-01 01:00:00', '2016-01-01 02:00:00']).indexer_between_time('01:00', '02:00', tz='Europe/Paris') TypeError: indexer_between_time() got an unexpected keyword argument 'tz'

10,929

pandas-dev/pandas

pandas-dev__pandas-12058

1945eed731a9d8fdb9a21837b326c42f8771def7

diff --git a/doc/source/whatsnew/v0.18.0.txt b/doc/source/whatsnew/v0.18.0.txt --- a/doc/source/whatsnew/v0.18.0.txt +++ b/doc/source/whatsnew/v0.18.0.txt @@ -445,7 +445,7 @@ Bug Fixes - Accept unicode in ``Timedelta`` constructor (:issue:`11995`) - Bug in value label reading for ``StataReader`` when reading incrementally (:issue:`12014`) - Bug in vectorized ``DateOffset`` when ``n`` parameter is ``0`` (:issue:`11370`) - +- Compat for numpy 1.11 w.r.t. ``NaT`` comparison changes (:issue:`12049`) diff --git a/pandas/core/common.py b/pandas/core/common.py --- a/pandas/core/common.py +++ b/pandas/core/common.py @@ -379,12 +379,13 @@ def array_equivalent(left, right, strict_nan=False): """ left, right = np.asarray(left), np.asarray(right) + + # shape compat if left.shape != right.shape: return False # Object arrays can contain None, NaN and NaT. - if (issubclass(left.dtype.type, np.object_) or - issubclass(right.dtype.type, np.object_)): + if is_object_dtype(left) or is_object_dtype(right): if not strict_nan: # pd.isnull considers NaN and None to be equivalent. @@ -405,13 +406,21 @@ def array_equivalent(left, right, strict_nan=False): return True # NaNs can occur in float and complex arrays. - if issubclass(left.dtype.type, (np.floating, np.complexfloating)): + if is_float_dtype(left) or is_complex_dtype(left): return ((left == right) | (np.isnan(left) & np.isnan(right))).all() # numpy will will not allow this type of datetimelike vs integer comparison elif is_datetimelike_v_numeric(left, right): return False + # M8/m8 + elif needs_i8_conversion(left) and needs_i8_conversion(right): + if not is_dtype_equal(left.dtype, right.dtype): + return False + + left = left.view('i8') + right = right.view('i8') + # NaNs cannot occur otherwise. return np.array_equal(left, right)

BLD: numpy master changes breaking (24 hrs ago) good build: https://travis-ci.org/pydata/pandas/jobs/102356098: 1.11.0.dev0+51d2ecd (1 hr ago) breaking lots of things: https://travis-ci.org/pydata/pandas/jobs/102596904: 1.11.0.dev0+aa6335c @shoyer IIRC a couple of your PR's were merged in the last day. here's an example: ``` ====================================================================== FAIL: test_coercion_with_setitem_and_series (pandas.tests.test_indexing.TestSeriesNoneCoercion) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/test_indexing.py", line 5247, in test_coercion_with_setitem_and_series expected_series.values, strict_nan=True) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 866, in assert_numpy_array_equal raise_assert_detail(obj, msg, left, right) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 825, in raise_assert_detail raise AssertionError(msg) AssertionError: numpy array are different numpy array values are different (33.33333 %) [left]: [NaT, 2000-01-02T00:00:00.000000000+0000, 2000-01-03T00:00:00.000000000+0000] [right]: [NaT, 2000-01-02T00:00:00.000000000+0000, 2000-01-03T00:00:00.000000000+0000] ```

my guess is that we now need to compare `M8[ns]` by `.view('i8')` then comparing as the `NaT` will then compare equal. Yep, if this is datetime64 related it's my fault :). I though we were already using .view('i8') before making comparisons but I guess I was wrong. If so (especially if this breaks user facing stuff) we may need to hold off on these numpy fixes for a little longer (deprecation cycle?). Sigh... On Fri, Jan 15, 2016 at 7:27 AM, Jeff Reback notifications@github.com wrote: > ## my guess is that we now need to compare `M8[ns]` by `.view('i8')` then comparing as the `NaT` will then compare equal. > > Reply to this email directly or view it on GitHub: > https://github.com/pydata/pandas/issues/12049#issuecomment-171991205 we can adapt. do you want me to open an issue on numpy? ``` In [1]: np.__version__ Out[1]: '1.10.2' In [5]: arr = np.array([np.nan]) In [6]: np.array_equal(arr,arr) Out[6]: False In [7]: arr = np.array([np.datetime64('NaT')]) In [8]: np.array_equal(arr,arr) Out[8]: True ``` ok so your change makes sense ``` In [1]: np.__version__ Out[1]: '1.11.0.dev0+aa6335c' In [2]: In [5]: arr = np.array([np.nan]) In [7]: arr = np.array([np.nan]) In [8]: np.array_equal(arr,arr) Out[8]: False In [9]: arr = np.array([np.datetime64('NaT')]) In [10]: np.array_equal(arr,arr) Out[10]: False ``` Actually, looks like we might just be able to drop our `assert_numpy_array_equivalent` entirely in favor of numpy's `assert_array_equal`, which has supported NaN equality since at least August 2011 (numpy 1.7, I think): https://github.com/numpy/numpy/commit/67ece6bdd2b35d011893e78154dbff6ab51c7d35 Unfortunately, this bug exists in our `array_equivalent` utility function, which we use for the `equals` method. This means that with this change datetime64 equality checks involving NaT will be broken. As much as I would love to just roll out the NumPy fix, doing it like this will assuredly result in unhappy users and unnecessary aggravation when sanity checks and test suites fail. To fix this, I propose: - We roll back the NaT comparison fix in NumPy for now, issuing a deprecation warning instead for a numpy release or two. - We add the fix to `array_equivalent` in pandas, doing the appropriate cast to int64 to avoid needing to catch the deprecation warning (which can have performance consequences). yep, prob need a special case for `M8/m8` (alternatively we could have a conditional check on the numpy version). Ok, lmk if you need anything on the numpy side.

2016-01-15T20:59:07Z

[]

Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/test_indexing.py", line 5247, in test_coercion_with_setitem_and_series expected_series.values, strict_nan=True) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 866, in assert_numpy_array_equal raise_assert_detail(obj, msg, left, right) File "/home/travis/build/pydata/pandas/pandas/util/testing.py", line 825, in raise_assert_detail raise AssertionError(msg) AssertionError: numpy array are different

10,930

pandas-dev/pandas

pandas-dev__pandas-13188

3944a369265f27268d1b3867a161e97f9c63cd62

diff --git a/pandas/tools/plotting.py b/pandas/tools/plotting.py --- a/pandas/tools/plotting.py +++ b/pandas/tools/plotting.py @@ -3353,7 +3353,8 @@ def _subplots(naxes=None, sharex=False, sharey=False, squeeze=True, if sharex or sharey: warnings.warn("When passing multiple axes, sharex and sharey " "are ignored. These settings must be specified " - "when creating axes", UserWarning) + "when creating axes", UserWarning, + stacklevel=4) if len(ax) == naxes: fig = ax[0].get_figure() return fig, ax @@ -3370,7 +3371,8 @@ def _subplots(naxes=None, sharex=False, sharey=False, squeeze=True, return fig, _flatten(ax) else: warnings.warn("To output multiple subplots, the figure containing " - "the passed axes is being cleared", UserWarning) + "the passed axes is being cleared", UserWarning, + stacklevel=4) fig.clear() nrows, ncols = _get_layout(naxes, layout=layout, layout_type=layout_type)

FAIL: test_scatter_matrix_axis (pandas.tests.test_graphics_others.TestDataFramePlots) in 0.18.1 with py27 When run alone, `test_scatter_matrix_axis` passes. When run together with the rest of `TestDataFramePlots`, it fails. I cannot test with py34 because matplotlib is not available on FreeBSD under py3k. #### Code Sample, a copy-pastable example if possible ``` % nosetests-2.7 pandas.tests.test_graphics_others:TestDataFramePlots.test_scatter_matrix_axis . ---------------------------------------------------------------------- Ran 1 test in 8.150s OK % nosetests-2.7 pandas.tests.test_graphics_others:TestDataFramePlots ..S/usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) /usr/local/lib/python2.7/site-packages/matplotlib/artist.py:221: MatplotlibDeprecationWarning: This has been deprecated in mpl 1.5, please use the axes property. A removal date has not been set. warnings.warn(_get_axes_msg, mplDeprecation, stacklevel=1) /usr/local/lib/python2.7/site-packages/pandas/tools/plotting.py:3369: UserWarning: To output multiple subplots, the figure containing the passed axes is being cleared "the passed axes is being cleared", UserWarning) ......F. ====================================================================== FAIL: test_scatter_matrix_axis (pandas.tests.test_graphics_others.TestDataFramePlots) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/local/lib/python2.7/site-packages/pandas/tests/test_graphics_others.py", line 431, in test_scatter_matrix_axis frame=df, range_padding=.1) File "/usr/local/lib/python2.7/contextlib.py", line 24, in __exit__ self.gen.next() File "/usr/local/lib/python2.7/site-packages/pandas/util/testing.py", line 2318, in assert_produces_warning % expected_warning.__name__) AssertionError: Did not see expected warning of class 'UserWarning'. ---------------------------------------------------------------------- Ran 11 tests in 74.896s FAILED (SKIP=1, failures=1) ``` #### Expected Output I expect all tests to pass. #### output of `pd.show_versions()` ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.11.final.0 python-bits: 64 OS: FreeBSD OS-release: 10.2-STABLE machine: amd64 processor: amd64 byteorder: little LC_ALL: None LANG: None pandas: 0.18.1 nose: 1.3.7 pip: 8.0.2 setuptools: 20.0 Cython: None numpy: 1.11.0 scipy: 0.16.1 statsmodels: None xarray: None IPython: None sphinx: None patsy: None dateutil: 2.5.0 pytz: 2016.4 blosc: None bottleneck: 1.0.0 tables: 3.2.2 numexpr: 2.5.2 matplotlib: 1.5.0 openpyxl: 2.3.5 xlrd: 0.9.3 xlwt: 0.7.5 xlsxwriter: 0.8.5 lxml: 3.5.0 bs4: 4.4.1 html5lib: 0.9999999 httplib2: None apiclient: None sqlalchemy: 0.7.10 pymysql: None psycopg2: None jinja2: 2.8 boto: 2.39.0 pandas_datareader: None ```

Thanks. Haven't been able to reproduce yet, but I'm going to clean up that testing module anyway to catch all those userwarnings and fix the matplotlib deprecation warning. Hopefully I'll figure out what's going wrong.

2016-05-15T17:11:08Z

[]

Traceback (most recent call last): File "/usr/local/lib/python2.7/site-packages/pandas/tests/test_graphics_others.py", line 431, in test_scatter_matrix_axis frame=df, range_padding=.1) File "/usr/local/lib/python2.7/contextlib.py", line 24, in __exit__ self.gen.next() File "/usr/local/lib/python2.7/site-packages/pandas/util/testing.py", line 2318, in assert_produces_warning % expected_warning.__name__) AssertionError: Did not see expected warning of class 'UserWarning'.

10,937

pandas-dev/pandas

pandas-dev__pandas-13641

20de2661c8eff66e465248cbe28062eae0e0e3bb

Test failure with matplotlib 1.5.2rc2 on Debian When running the test suite after the package was built, we get the following failure: ``` ERROR: test_plot (pandas.tests.test_graphics.TestDataFramePlots) ---------------------------------------------------------------------- Traceback (most recent call last): File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tests/test_graphics.py", line 1322, in test_plot df.plot.line(blarg=True) File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 3758, in line return self(kind='line', x=x, y=y, **kwds) [...] File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 1340, in _plot return ax.plot(*args, **kwds) File "/usr/lib/python2.7/dist-packages/matplotlib/__init__.py", line 1821, in inner return func(ax, *args, **kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_axes.py", line 1432, in plot for line in self._get_lines(*args, **kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 381, in _grab_next_args for seg in self._plot_args(remaining, kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 369, in _plot_args seg = func(x[:, j % ncx], y[:, j % ncy], kw, kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 276, in _makeline seg = mlines.Line2D(x, y, **kw) File "/usr/lib/python2.7/dist-packages/matplotlib/lines.py", line 380, in __init__ self.update(kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/artist.py", line 859, in update raise AttributeError('Unknown property %s' % k) AttributeError: Unknown property blarg ``` Full log [here](https://tests.reproducible-builds.org/debian/rbuild/testing/amd64/pandas_0.18.0+git114-g6c692ae-1.rbuild.log) (from the [Debian bug report](https://bugs.debian.org/827938)). `pandas.show_versions()` gives: ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.12.final.0 python-bits: 64 OS: Linux OS-release: 4.6.0-1-amd64 machine: x86_64 processor: byteorder: little LC_ALL: C LANG: C pandas: 0.18.0+git114-g6c692ae nose: 1.3.7 pip: None setuptools: 20.10.1 Cython: None numpy: 1.11.1rc1 scipy: 0.17.1 statsmodels: 0.6.1 xarray: None IPython: None sphinx: 1.4.4 patsy: 0.4.1 dateutil: 2.4.2 pytz: 2015.7 blosc: None bottleneck: None tables: 3.2.2 numexpr: 2.6.0 matplotlib: 1.5.2rc2 openpyxl: 2.3.0 xlrd: 1.0.0 xlwt: 0.7.5 xlsxwriter: None lxml: 3.6.0 bs4: 4.4.1 html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: None pymysql: None psycopg2: None jinja2: 2.8 boto: None pandas_datareader: None ```

Thanks for the report. Looks to be caused by mpl which raises `AttributeError` rather than `TypeError`. Should change the test on pandas side. - https://github.com/pydata/pandas/blob/master/pandas/tests/test_graphics.py#L1317 Appreciated to submit a PR. [line 1333](https://github.com/pydata/pandas/blob/master/pandas/tests/test_graphics.py#L1333) Shall this be compatible with matplotlib < 1.5.2rc? Then this is not that simple, since tm.assertRaises() only accepts a singe exception, and one then would need to create another mpl_ge_1_5_2 comparison specifically for this test. Is this really worth it? What is the rationale of the test? cc @tacaswell Just to be sure, the change from TypeError to AttributeError, was this on purpose? (it's raised in `set_lineprops`) I do not think that was intentional and I am not sure off the top of my head why this changed. The `Artist.update` method has raised `AttributeError` since 2004. So, I broke this: https://github.com/matplotlib/matplotlib/pull/6175 It looks like we used to have `Artist.set` which raised `TypeError` and `Artist.update` which raised `AttributeError` (because history). These two code-paths got merged in https://github.com/matplotlib/matplotlib/pull/5599 (also my fault) and we missed the API change in the the exceptions. Unfortunately I _just_ tagged 1.5.2 last weekend, but have not posted it to pypi or publicized it yet :disappointed:. I am inclined to just document this as an API change, but if you want to lobby for a 1.5.3 fixing this, this is the time to do it! I don't think it is that an important change, so leaving (and documenting) it is fine for me (for pandas it is only a test that is broken, and I think you will have a better idea of how big this change is for matploltib users). In any case, I think the AttributeError is more logical. @sinhrks @jorisvandenbossche For me the question here is still, what that test is for? can u add `_mpl_ge_1_5_2()` function like below, then: ``` if _mpl_ge_1_5_2(): with tm.assertRaises(AttributeError): ... else: with ... ``` - https://github.com/pydata/pandas/blob/master/pandas/tools/plotting.py#L129 > For me the question here is still, what that test is for? Well, it tests that if you pass a wrong keyword, that this raises an error, and is not swallowed somewhere in the implementation. So it's not a huge important test, but still useful (we actually have to many functions in pandas that silently swallow invalid arguments)

2016-07-13T15:01:57Z

[]

Traceback (most recent call last): File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tests/test_graphics.py", line 1322, in test_plot df.plot.line(blarg=True) File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 3758, in line return self(kind='line', x=x, y=y, **kwds) [...] File "/$BUILD/debian/tmp/usr/lib/python2.7/dist-packages/pandas/tools/plotting.py", line 1340, in _plot return ax.plot(*args, **kwds) File "/usr/lib/python2.7/dist-packages/matplotlib/__init__.py", line 1821, in inner return func(ax, *args, **kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_axes.py", line 1432, in plot for line in self._get_lines(*args, **kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 381, in _grab_next_args for seg in self._plot_args(remaining, kwargs): File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 369, in _plot_args seg = func(x[:, j % ncx], y[:, j % ncy], kw, kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/axes/_base.py", line 276, in _makeline seg = mlines.Line2D(x, y, **kw) File "/usr/lib/python2.7/dist-packages/matplotlib/lines.py", line 380, in __init__ self.update(kwargs) File "/usr/lib/python2.7/dist-packages/matplotlib/artist.py", line 859, in update raise AttributeError('Unknown property %s' % k) AttributeError: Unknown property blarg

10,954

pandas-dev/pandas

pandas-dev__pandas-14126

47a8e713430707afcfe76e7ca995902628d4bccf

diff --git a/pandas/util/print_versions.py b/pandas/util/print_versions.py --- a/pandas/util/print_versions.py +++ b/pandas/util/print_versions.py @@ -101,7 +101,10 @@ def show_versions(as_json=False): deps_blob = list() for (modname, ver_f) in deps: try: - mod = importlib.import_module(modname) + if modname in sys.modules: + mod = sys.modules[modname] + else: + mod = importlib.import_module(modname) ver = ver_f(mod) deps_blob.append((modname, ver)) except:

DataFrame.__repr__ raises TypeError after pd.show_versions() was run Maybe one of the imports in `show_versions` has unwanted side effects? ``` python >>> import pandas as pd >>> pd.DataFrame({'spam': range(10)}) spam 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 >>> pd.show_versions() INSTALLED VERSIONS ------------------ commit: None python: 2.7.12.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 42 Stepping 7, GenuineIntel byteorder: little LC_ALL: None LANG: None pandas: 0.18.1 nose: 1.3.7 pip: 8.1.2 setuptools: 24.0.3 Cython: 0.24.1 numpy: 1.11.1 scipy: 0.18.0rc2 statsmodels: None xarray: None IPython: 5.0.0 sphinx: 1.4.5 patsy: 0.4.1 dateutil: 2.5.3 pytz: 2016.6.1 blosc: None bottleneck: 1.1.0 tables: None numexpr: 2.6.0 matplotlib: 1.5.1 openpyxl: 2.3.5 xlrd: 1.0.0 xlwt: None xlsxwriter: None lxml: 3.6.0 bs4: None html5lib: 0.999999999 httplib2: 0.9.2 apiclient: None sqlalchemy: 1.0.14 pymysql: None psycopg2: 2.6.2 (dt dec pq3 ext lo64) jinja2: 2.8 boto: None pandas_datareader: None >>> pd.DataFrame({'spam': range(10)}) Traceback (most recent call last): File "<pyshell#3>", line 1, in <module> pd.DataFrame({'spam': range(10)}) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 67, in __repr__ return str(self) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 47, in __str__ return self.__bytes__() File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 59, in __bytes__ return self.__unicode__().encode(encoding, 'replace') File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 535, in __unicode__ line_width=width, show_dimensions=show_dimensions) File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 1488, in to_string formatter.to_string() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 549, in to_string strcols = self._to_str_columns() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 467, in _to_str_columns str_index = self._get_formatted_index(frame) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 746, in _get_formatted_index fmt_index = [index.format(name=show_index_names, formatter=fmt)] File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1462, in format return self._format_with_header(header, **kwargs) File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1486, in _format_with_header result = _trim_front(format_array(values, None, justify='left')) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2007, in format_array return fmt_obj.get_result() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2027, in get_result return _make_fixed_width(fmt_values, self.justify) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2394, in _make_fixed_width max_len = np.max([adj.len(x) for x in strings]) File "C:\Program Files\Python27\lib\site-packages\numpy\core\fromnumeric.py", line 2293, in amax out=out, **kwargs) File "C:\Program Files\Python27\lib\site-packages\numpy\core\_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) TypeError: an integer is required ```

I cannot reproduce this using Windows and python 2.7. Could you try to debug this? To see where the error is coming from (seems there is something wrong with the `np.max([adj.len(x) for x in strings])`). Or eg create an isolated environment with only required dependencies to see of the problem occurs there as well (using conda or virtualenv). Tracked it to [this import](https://github.com/pydata/pandas/commit/b4e2d34edcbc404f6c90f76b67bcc5fe26f0945f#diff-24212510f4a09e0461c2b6754d34626dL103) of `numpy`, which according to [the docs](https://docs.python.org/2/library/imp.html#imp.load_module) does a `reload()`. Indeed `numpy` (at least on my machines) seems to dislike being reloaded: ``` python >>> import numpy as np >>> np.max([42]) 42 >>> reload(np) <module 'numpy' from 'C:\Program Files\Python27\lib\site-packages\numpy\__init__.pyc'> >>> np.max([42]) Traceback (most recent call last): File "<pyshell#3>", line 1, in <module> np.max([42]) File "C:\Program Files\Python27\lib\site-packages\numpy\core\fromnumeric.py", line 2293, in amax out=out, **kwargs) File "C:\Program Files\Python27\lib\site-packages\numpy\core\_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) TypeError: an integer is required ``` The line was changed in b4e2d34edcbc404f6c90f76b67bcc5fe26f0945f, so I guess that should fix this for the next version (though I am still curious if others have this `numpy` issue with `reload`). I can't reproduce it using NumPy 1.10.4 and 1.11.1. Can you report it to NumPy? Seems to be a problem with the [binaries here](http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy) (does not occur with the PyPI wheels): I'll report to the packager, closing. The numpy issue: numpy/numpy#7844 @xflr6 Thanks for tracking it down! I think there is a real pandas bug here. The bug is that `show_versions` calls [`importlib.import_module`](https://github.com/pydata/pandas/blob/b4e2d34edcbc404f6c90f76b67bcc5fe26f0945f/pandas/util/print_versions.py#L102), and apparently -- this is not documented anywhere, and may vary between py2 and py3 -- `import_module` may reload modules. `show_versions` should not be reloading all these modules. I'd suggest replacing that line with something like ``` python if modname in sys.modules: mod = sys.modules[modname] else: mod = importlib.import_module(modname) ``` @njsmith but seems numpy is not robust to being reloaded. @jreback I suspect a lot of modules are not robust against reloading. For instance if you define a class in `foo.py`, use it in `bar.py`, instanciate `a = foo.MyClass()` in another module, then use `isinstance(a, foo.MyClass)` that statement will fail if `foo` is reloaded. I suspect what is wanted in many cases is a simple `import`, but I haven't checked. Numpy also uses `load_module` in a few places that should probably be audited. @njsmith The reload property is documented in the `imp` module documentation. It always happens for existing modules. Note that nonexisting modules get created... ok this should be easy to fix then AFAICT, `load_module` is useful when you need to use a module that is not installed and not located in the current directory. For instance, during the numpy install process. If numpy is installed you should be able to simply import it. Maybe `__import__`? Or `importlib.import_module` Pandas actually uses `importlib.import_module`, which isn't documented to reload, but I guess it must eventually call `load_module` because otherwise we wouldn't have this problem. (I haven't tried tracing the details, and `importlib` has completely different implementations on different versions of python, so that's something to watch out for if anyone wants to figure out exactly what's happening).

2016-08-31T10:35:26Z

[]

Traceback (most recent call last): File "<pyshell#3>", line 1, in <module> pd.DataFrame({'spam': range(10)}) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 67, in __repr__ return str(self) File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 47, in __str__ return self.__bytes__() File "C:\Program Files\Python27\lib\site-packages\pandas\core\base.py", line 59, in __bytes__ return self.__unicode__().encode(encoding, 'replace') File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 535, in __unicode__ line_width=width, show_dimensions=show_dimensions) File "C:\Program Files\Python27\lib\site-packages\pandas\core\frame.py", line 1488, in to_string formatter.to_string() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 549, in to_string strcols = self._to_str_columns() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 467, in _to_str_columns str_index = self._get_formatted_index(frame) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 746, in _get_formatted_index fmt_index = [index.format(name=show_index_names, formatter=fmt)] File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1462, in format return self._format_with_header(header, **kwargs) File "C:\Program Files\Python27\lib\site-packages\pandas\indexes\base.py", line 1486, in _format_with_header result = _trim_front(format_array(values, None, justify='left')) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2007, in format_array return fmt_obj.get_result() File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2027, in get_result return _make_fixed_width(fmt_values, self.justify) File "C:\Program Files\Python27\lib\site-packages\pandas\formats\format.py", line 2394, in _make_fixed_width max_len = np.max([adj.len(x) for x in strings]) File "C:\Program Files\Python27\lib\site-packages\numpy\core\fromnumeric.py", line 2293, in amax out=out, **kwargs) File "C:\Program Files\Python27\lib\site-packages\numpy\core\_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) TypeError: an integer is required

11,001

pandas-dev/pandas

pandas-dev__pandas-14208

5e2f9da6e8e713bd89cfe8760e63583ea7d29879

diff --git a/doc/source/basics.rst b/doc/source/basics.rst --- a/doc/source/basics.rst +++ b/doc/source/basics.rst @@ -188,6 +188,32 @@ And similarly for ``axis="items"`` and ``axis="minor"``. match the broadcasting behavior of Panel. Though it would require a transition period so users can change their code... +Series and Index also support the :func:`divmod` builtin. This function takes +the floor division and modulo operation at the same time returning a two-tuple +of the same type as the left hand side. For example: + +.. ipython:: python + + s = pd.Series(np.arange(10)) + s + div, rem = divmod(s, 3) + div + rem + + idx = pd.Index(np.arange(10)) + idx + div, rem = divmod(idx, 3) + div + rem + +We can also do elementwise :func:`divmod`: + +.. ipython:: python + + div, rem = divmod(s, [2, 2, 3, 3, 4, 4, 5, 5, 6, 6]) + div + rem + Missing data / operations with fill values ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ diff --git a/doc/source/whatsnew/v0.19.0.txt b/doc/source/whatsnew/v0.19.0.txt --- a/doc/source/whatsnew/v0.19.0.txt +++ b/doc/source/whatsnew/v0.19.0.txt @@ -1328,6 +1328,9 @@ Other API Changes - ``pd.read_csv()`` in the C engine will now issue a ``ParserWarning`` or raise a ``ValueError`` when ``sep`` encoded is more than one character long (:issue:`14065`) - ``DataFrame.values`` will now return ``float64`` with a ``DataFrame`` of mixed ``int64`` and ``uint64`` dtypes, conforming to ``np.find_common_type`` (:issue:`10364`, :issue:`13917`) - ``pd.read_stata()`` can now handle some format 111 files, which are produced by SAS when generating Stata dta files (:issue:`11526`) +- ``Series`` and ``Index`` now support ``divmod`` which will return a tuple of + series or indices. This behaves like a standard binary operator with regards + to broadcasting rules (:issue:`14208`). .. _whatsnew_0190.deprecations: diff --git a/pandas/core/ops.py b/pandas/core/ops.py --- a/pandas/core/ops.py +++ b/pandas/core/ops.py @@ -39,7 +39,8 @@ def _create_methods(arith_method, comp_method, bool_method, - use_numexpr, special=False, default_axis='columns'): + use_numexpr, special=False, default_axis='columns', + have_divmod=False): # creates actual methods based upon arithmetic, comp and bool method # constructors. @@ -127,6 +128,15 @@ def names(x): names('ror_'), op('|')), rxor=bool_method(lambda x, y: operator.xor(y, x), names('rxor'), op('^')))) + if have_divmod: + # divmod doesn't have an op that is supported by numexpr + new_methods['divmod'] = arith_method( + divmod, + names('divmod'), + None, + default_axis=default_axis, + construct_result=_construct_divmod_result, + ) new_methods = dict((names(k), v) for k, v in new_methods.items()) return new_methods @@ -156,7 +166,7 @@ def add_methods(cls, new_methods, force, select, exclude): def add_special_arithmetic_methods(cls, arith_method=None, comp_method=None, bool_method=None, use_numexpr=True, force=False, select=None, - exclude=None): + exclude=None, have_divmod=False): """ Adds the full suite of special arithmetic methods (``__add__``, ``__sub__``, etc.) to the class. @@ -177,6 +187,9 @@ def add_special_arithmetic_methods(cls, arith_method=None, if passed, only sets functions with names in select exclude : iterable of strings (optional) if passed, will not set functions with names in exclude + have_divmod : bool, (optional) + should a divmod method be added? this method is special because it + returns a tuple of cls instead of a single element of type cls """ # in frame, special methods have default_axis = None, comp methods use @@ -184,7 +197,7 @@ def add_special_arithmetic_methods(cls, arith_method=None, new_methods = _create_methods(arith_method, comp_method, bool_method, use_numexpr, default_axis=None, - special=True) + special=True, have_divmod=have_divmod) # inplace operators (I feel like these should get passed an `inplace=True` # or just be removed @@ -618,8 +631,22 @@ def _align_method_SERIES(left, right, align_asobject=False): return left, right +def _construct_result(left, result, index, name, dtype): + return left._constructor(result, index=index, name=name, dtype=dtype) + + +def _construct_divmod_result(left, result, index, name, dtype): + """divmod returns a tuple of like indexed series instead of a single series. + """ + constructor = left._constructor + return ( + constructor(result[0], index=index, name=name, dtype=dtype), + constructor(result[1], index=index, name=name, dtype=dtype), + ) + + def _arith_method_SERIES(op, name, str_rep, fill_zeros=None, default_axis=None, - **eval_kwargs): + construct_result=_construct_result, **eval_kwargs): """ Wrapper function for Series arithmetic operations, to avoid code duplication. @@ -692,8 +719,14 @@ def wrapper(left, right, name=name, na_op=na_op): lvalues = lvalues.values result = wrap_results(safe_na_op(lvalues, rvalues)) - return left._constructor(result, index=left.index, - name=name, dtype=dtype) + return construct_result( + left, + result, + index=left.index, + name=name, + dtype=dtype, + ) + return wrapper @@ -933,6 +966,10 @@ def wrapper(self, other): 'desc': 'Integer division', 'reversed': False, 'reverse': 'rfloordiv'}, + 'divmod': {'op': 'divmod', + 'desc': 'Integer division and modulo', + 'reversed': False, + 'reverse': None}, 'eq': {'op': '==', 'desc': 'Equal to', @@ -1033,7 +1070,8 @@ def flex_wrapper(self, other, level=None, fill_value=None, axis=0): series_special_funcs = dict(arith_method=_arith_method_SERIES, comp_method=_comp_method_SERIES, - bool_method=_bool_method_SERIES) + bool_method=_bool_method_SERIES, + have_divmod=True) _arith_doc_FRAME = """ Binary operator %s with support to substitute a fill_value for missing data in diff --git a/pandas/indexes/base.py b/pandas/indexes/base.py --- a/pandas/indexes/base.py +++ b/pandas/indexes/base.py @@ -3426,7 +3426,7 @@ def _validate_for_numeric_binop(self, other, op, opstr): def _add_numeric_methods_binary(cls): """ add in numeric methods """ - def _make_evaluate_binop(op, opstr, reversed=False): + def _make_evaluate_binop(op, opstr, reversed=False, constructor=Index): def _evaluate_numeric_binop(self, other): from pandas.tseries.offsets import DateOffset @@ -3448,7 +3448,7 @@ def _evaluate_numeric_binop(self, other): attrs = self._maybe_update_attributes(attrs) with np.errstate(all='ignore'): result = op(values, other) - return Index(result, **attrs) + return constructor(result, **attrs) return _evaluate_numeric_binop @@ -3478,6 +3478,15 @@ def _evaluate_numeric_binop(self, other): cls.__rdiv__ = _make_evaluate_binop( operator.div, '__div__', reversed=True) + cls.__divmod__ = _make_evaluate_binop( + divmod, + '__divmod__', + constructor=lambda result, **attrs: ( + Index(result[0], **attrs), + Index(result[1], **attrs), + ), + ) + @classmethod def _add_numeric_methods_unary(cls): """ add in numeric unary methods """

Regression: divmod(my_series, some_integer) no longer works since version 0.13. In Pandas version 0.12, you could apply the Python built-in `divmod` function to a `Series` and an integer: ``` Enthought Canopy Python 2.7.6 | 64-bit | (default, Jun 4 2014, 16:42:26) [GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import pandas as pd >>> pd.__version__ '0.12.0' >>> divmod(pd.Series(range(4)), 2) (0 0 1 0 2 1 3 1 dtype: int64, 0 0 1 1 2 0 3 1 dtype: int64) >>> ``` With version >= 0.13, it appears that this usage is no longer supported: ``` Enthought Canopy Python 2.7.6 | 64-bit | (default, Jun 4 2014, 16:42:26) [GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import pandas as pd >>> pd.__version__ '0.14.1' >>> divmod(pd.Series(range(4)), 2) Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: unsupported operand type(s) for divmod(): 'Series' and 'int' ``` Was this change intentional? Some context: I was using this to read a climate datafile that had a 4-digit column holding combined month and day values. The original code looked something like: `month, day = divmod(df['MODA'], 100)`, but broke after upgrading to version 0.14.

`__divmod__` was not included when Series was refactored in 0.13. You can simply do: `s // 2, s % 2` if you want to get the same results If you would like to add the method, a PR would be accepted. I don't think was included as no tests for this (and to be honest never seen it used; its more of a 'functional' way of working with objects)

2016-09-12T18:58:01Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: unsupported operand type(s) for divmod(): 'Series' and 'int'

11,005

pandas-dev/pandas

pandas-dev__pandas-14225

e8357a15cd61ff698cbd3d57904133c586a8ed8b

diff --git a/doc/source/whatsnew/v0.19.0.txt b/doc/source/whatsnew/v0.19.0.txt --- a/doc/source/whatsnew/v0.19.0.txt +++ b/doc/source/whatsnew/v0.19.0.txt @@ -1572,3 +1572,4 @@ Bug Fixes - Bug in ``eval()`` where the ``resolvers`` argument would not accept a list (:issue:`14095`) - Bugs in ``stack``, ``get_dummies``, ``make_axis_dummies`` which don't preserve categorical dtypes in (multi)indexes (:issue:`13854`) - ``PeridIndex`` can now accept ``list`` and ``array`` which contains ``pd.NaT`` (:issue:`13430`) +- Bug in ``df.groupby`` where ``.median()`` returns arbitrary values if grouped dataframe contains empty bins (:issue:`13629`) diff --git a/pandas/algos.pyx b/pandas/algos.pyx --- a/pandas/algos.pyx +++ b/pandas/algos.pyx @@ -992,7 +992,7 @@ def is_lexsorted(list list_of_arrays): def groupby_indices(dict ids, ndarray[int64_t] labels, ndarray[int64_t] counts): """ - turn group_labels output into a combined indexer maping the labels to + turn group_labels output into a combined indexer mapping the labels to indexers Parameters @@ -1313,6 +1313,9 @@ cdef inline float64_t _median_linear(float64_t* a, int n): cdef float64_t result cdef float64_t* tmp + if n == 0: + return NaN + # count NAs for i in range(n): if a[i] != a[i]: diff --git a/pandas/core/groupby.py b/pandas/core/groupby.py --- a/pandas/core/groupby.py +++ b/pandas/core/groupby.py @@ -4424,12 +4424,13 @@ def _reorder_by_uniques(uniques, labels): def _groupby_indices(values): if is_categorical_dtype(values): - # we have a categorical, so we can do quite a bit # bit better than factorizing again reverse = dict(enumerate(values.categories)) codes = values.codes.astype('int64') - _, counts = _hash.value_count_int64(codes, False) + + mask = 0 <= codes + counts = np.bincount(codes[mask], minlength=values.categories.size) else: reverse, codes, counts = _algos.group_labels( _values_from_object(_ensure_object(values)))

BUG: Binned groupby median function calculates median on empty bins and outputs random numbers #### Code Sample, a copy-pastable example if possible ``` import pandas as pd d = pd.DataFrame([1,2,5,6,9,3,6,5,9,7,11,36,4,7,8,25,8,24]) b = [0,5,10,15,20,25,30,35,40,45,50,55] g = d.groupby(pd.cut(d[0],b)) print g.mean() print g.median() print g.get_group('(0, 5]').median() print g.get_group('(40, 45]').median() ``` #### Expected Output ``` 0 0 (0, 5] 3.333333 (5, 10] 7.500000 (10, 15] 11.000000 (15, 20] NaN (20, 25] 24.500000 (25, 30] NaN (30, 35] NaN (35, 40] 36.000000 (40, 45] NaN (45, 50] NaN (50, 55] NaN 0 0 (0, 5] 3.5 (5, 10] 7.5 (10, 15] 11.0 (15, 20] 18.0 (20, 25] 24.5 (25, 30] 30.5 (30, 35] 30.5 (35, 40] 36.0 (40, 45] 18.0 (45, 50] 18.0 (50, 55] 18.0 0 3.5 dtype: float64 Traceback (most recent call last): File "<ipython-input-9-0663486889da>", line 1, in <module> runfile('C:/PythonDir/test04.py', wdir='C:/PythonDir') File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 714, in runfile execfile(filename, namespace) File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 74, in execfile exec(compile(scripttext, filename, 'exec'), glob, loc) File "C:/PythonDir/test04.py", line 20, in <module> print g.get_group('(40, 45]').median() File "C:\Anaconda2\lib\site-packages\pandas\core\groupby.py", line 587, in get_group raise KeyError(name) KeyError: '(40, 45]' ``` This example shows how the median-function of the groupby object outputs a random number instead of NaN like the mean-function does when a bin is empty. Directly trying to call that bin with its key leads to an error since it doesn't exist, yet the full median output suggests it does exist and that the value might even be meaningful (like in the (15, 20] bin or the (30, 35] bin). The wrong numbers that are returned can change randomly, another possible output using the same code might look like this: ``` (0, 5] 3.500000e+00 (5, 10] 7.500000e+00 (10, 15] 1.100000e+01 (15, 20] 1.800000e+01 (20, 25] 2.450000e+01 (25, 30] 3.050000e+01 (30, 35] 3.050000e+01 (35, 40] 3.600000e+01 (40, 45] 4.927210e+165 (45, 50] 4.927210e+165 (50, 55] 4.927210e+165 ``` #### output of `pd.show_versions()` `pandas: 0.18.1`

@Khris777 Thanks for reporting! As a workaround for now, you can do: ``` In [11]: g.agg(lambda x: x.median()) Out[11]: 0 0 (0, 5] 3.5 (5, 10] 7.5 (10, 15] 11.0 (15, 20] NaN (20, 25] 24.5 (25, 30] NaN (30, 35] NaN (35, 40] 36.0 (40, 45] NaN (45, 50] NaN (50, 55] NaN ``` First time contributor, thought I'd take a look into this one. Do you think there's a more logical response than raising a KeyError to `g.get_group('(40, 45]')` ? get_group with no additional arguments is supposed to return a subset of the original dataframe with values that fall within the specified interval. If there are no values in the interval (40,45] in the original dataframe, there's no way to slice that up into a sensible response. Empty dataframe? ATM, internval types are actual string reprs (and not a distinct dtype), so yes, `g.get_group('(40, 45)')` should be a `KeyError`, just like any other indexing operation.

2016-09-15T01:57:41Z

[]

Traceback (most recent call last): File "<ipython-input-9-0663486889da>", line 1, in <module> runfile('C:/PythonDir/test04.py', wdir='C:/PythonDir') File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 714, in runfile execfile(filename, namespace) File "C:\Anaconda2\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 74, in execfile exec(compile(scripttext, filename, 'exec'), glob, loc) File "C:/PythonDir/test04.py", line 20, in <module> print g.get_group('(40, 45]').median() File "C:\Anaconda2\lib\site-packages\pandas\core\groupby.py", line 587, in get_group raise KeyError(name) KeyError: '(40, 45]'

11,007

pandas-dev/pandas

pandas-dev__pandas-14329

6dcc23862b6b60ce2a67436b4a278fbe4c05490f

diff --git a/pandas/core/frame.py b/pandas/core/frame.py --- a/pandas/core/frame.py +++ b/pandas/core/frame.py @@ -1635,7 +1635,8 @@ def to_latex(self, buf=None, columns=None, col_space=None, header=True, When set to False prevents from escaping latex special characters in column names. encoding : str, default None - Default encoding is ascii in Python 2 and utf-8 in Python 3 + A string representing the encoding to use in the output file, + defaults to 'ascii' on Python 2 and 'utf-8' on Python 3. decimal : string, default '.' Character recognized as decimal separator, e.g. ',' in Europe diff --git a/pandas/formats/format.py b/pandas/formats/format.py --- a/pandas/formats/format.py +++ b/pandas/formats/format.py @@ -654,6 +654,9 @@ def to_latex(self, column_format=None, longtable=False, encoding=None): latex_renderer = LatexFormatter(self, column_format=column_format, longtable=longtable) + if encoding is None: + encoding = 'ascii' if compat.PY2 else 'utf-8' + if hasattr(self.buf, 'write'): latex_renderer.write_result(self.buf) elif isinstance(self.buf, compat.string_types):

TST: 3.5 c-locale https://travis-ci.org/pydata/pandas/jobs/161159736 xref #14114, #12337 ``` ====================================================================== ERROR: test_to_latex_filename (pandas.tests.formats.test_format.TestDataFrameFormatting) ---------------------------------------------------------------------- Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/formats/test_format.py", line 2825, in test_to_latex_filename df.to_latex(path) File "/home/travis/build/pydata/pandas/pandas/core/frame.py", line 1661, in to_latex encoding=encoding) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 662, in to_latex latex_renderer.write_result(f) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 906, in write_result buf.write(' & '.join(crow)) UnicodeEncodeError: 'ascii' codec can't encode character '\xdf' in position 7: ordinal not in range(128) ---------------------------------------------------------------------- Ran 10273 tests in 375.734s ``` cc @nbonnotte

cc @nbonnotte

2016-10-01T17:34:21Z

[]

Traceback (most recent call last): File "/home/travis/build/pydata/pandas/pandas/tests/formats/test_format.py", line 2825, in test_to_latex_filename df.to_latex(path) File "/home/travis/build/pydata/pandas/pandas/core/frame.py", line 1661, in to_latex encoding=encoding) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 662, in to_latex latex_renderer.write_result(f) File "/home/travis/build/pydata/pandas/pandas/formats/format.py", line 906, in write_result buf.write(' & '.join(crow)) UnicodeEncodeError: 'ascii' codec can't encode character '\xdf' in position 7: ordinal not in range(128)

11,016

pandas-dev/pandas

pandas-dev__pandas-14853

b6de920d8c3c2becc46b4fe233e9f388947554f2

BUG: Passing ambiguous ndarray[datetime64[ns]] to DatetimeIndex constructor can cause ValueError with wrong offset if you give infer_freq 5 consecutive weekdays, it'll come back with 'D' as its inferred frequency. But if your actual frequency is `BDay`, then, when DatetimeIndex checks that the frequency matches, 'B' != 'D'. (note that verify_integrity=False skips this). This leads to a more general issue about `infer_freq` with ambiguous cases. I think it makes the most sense to move these sorts of checks to a method on offset that takes a frequence and an Index or ndarray, and determines whether it is compatible. This matters because you can hit some edge cases when you pass freq and also datetime64[ns] to the DatetimeIndex constructor and more generally because comparing freqstr is probably not the best way to go about checking whether a frequency matches. Default implementation could be: ``` python def is_compatible(freqstr, arr=None): return freqstr == self.freqstr ``` and then bday could do something like (and this is totally psuedocode) ``` python def is_compatible(freqstr, arr=None): if freqstr == self.freqstr: return True if arr is not None and len(arr) <= 5: if freqstr == 'D': # or other compatibles that ensure it's consecutive return all(is_weekday(date) for date in arr) ``` This gets more complicated with multiplied offsets, but I think it's worth considering. ``` python from datetime import datetime import pandas as pd dates = [datetime(2013, 10, 7), datetime(2013, 10, 8), datetime(2013, 10, 9)] ind = pd.DatetimeIndex(dates, freq=pd.tseries.frequencies.BDay()) ind2 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay(), verify_integrity=False) ind3 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay()) ``` produces this Traceback: ``` Traceback (most recent call last): File "test2.py", line 8, in <module> ind3 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay()) File "../pandas/tseries/index.py", line 280, in __new__ raise ValueError('Dates do not conform to passed ' ValueError: Dates do not conform to passed frequency ``` cc @cancan101 - this is what we need to deal with in adding your offsets. I believe that every other offset can be returned from infer_freq, so these offsets would be different and therefore could _never_ pass integrity checks. So either we'd need to change infer_freq and/or define some kind of is_compatible method that intelligently covers all the ways in which the frequency could be something different than its freqstr.

Pretty simple test case: ``` python In [88]: infer_freq([datetime(2013, 10, 7), datetime(2013, 10, 8), datetime(2013, 10, 9)]) Out[88]: 'D' ``` which I agree is incorrect since a frequency of 'BD' cannot be ruled out. @wesm ? no that's not incorrect - it's reasonable and valid as the freq for the sequence, right? It might be ambiguous, but all that matters is inferring some frequency. So either it should return a list of possibilities, or we pass the buck to offsets to handle the ambiguity. Okay. Agreed, "incorrect" is not the right word. That being said, what exactly is the spec for infer_freq? There is not quite a total ordering of frequencies by specificity, but in general, should infer_freq return the most specific or the more general frequency? @cancan101 it obviously is the most general, it is used quite extensively interally to lazily evaluate frequency when its not already assigned Perhaps `infer_freq` should have an option not to guess when there is any potential for ambiguity. This might be the default behavior, then. I don't think anyone will shed tears if a length-3 (or 5) array case like you describe gets inferred as no frequency with the change. @wesm or at the very least could skip frequency inference when small.

2016-12-10T21:23:26Z

[]

Traceback (most recent call last): File "test2.py", line 8, in <module> ind3 = pd.DatetimeIndex(ind.values, freq=pd.tseries.frequencies.BDay()) File "../pandas/tseries/index.py", line 280, in __new__ raise ValueError('Dates do not conform to passed ' ValueError: Dates do not conform to passed frequency

11,066

pandas-dev/pandas

pandas-dev__pandas-14884

3ba2cff9c55cd16b172f9feb09da551990753f3b

Assigning datetime array to column fails with OutOfBoundsDatetime when having NaT and other unit as [ns] Assigning an array with datetime64[ns] values including a NaT just works: ``` In [85]: a = np.array([1, 'nat'], dtype='datetime64[ns]') In [86]: pd.Series(a) Out[86]: 0 1970-01-01 00:00:00.000000001 1 NaT dtype: datetime64[ns] In [88]: df = pd.Series(a).to_frame() In [89]: df['new'] = a ``` But when having an array with another date unit, converting it to a Series still works, but assigning it directly to a column not anymore, resulting in a OutOfBoundsDatetime error: ``` In [90]: a = np.array([1, 'nat'], dtype='datetime64[s]') In [91]: pd.Series(a) Out[91]: 0 1970-01-01 00:00:01 1 NaT dtype: datetime64[ns] In [92]: df['new'] = a Traceback (most recent call last): ... File "tslib.pyx", line 1720, in pandas.tslib.cast_to_nanoseconds (pandas\tslib.c:27435) File "tslib.pyx", line 1023, in pandas.tslib._check_dts_bounds (pandas\tslib.c:18102) OutOfBoundsDatetime: Out of bounds nanosecond timestamp: 292277026596-12-03 08:29:52 ``` If you first convert it to a series, it does work. Also if the `NaT` is not present: ``` In [93]: df['new'] = pd.Series(a) In [94]: a = np.array([1, 2], dtype='datetime64[s]') In [95]: df['new'] = a ```

yep, this conversion is amazingly complex. marking as a bug

2016-12-15T06:06:56Z

[]

Traceback (most recent call last): ... File "tslib.pyx", line 1720, in pandas.tslib.cast_to_nanoseconds (pandas\tslib.c:27435) File "tslib.pyx", line 1023, in pandas.tslib._check_dts_bounds (pandas\tslib.c:18102) OutOfBoundsDatetime: Out of bounds nanosecond timestamp: 292277026596-12-03 08:29:52

11,069

pandas-dev/pandas

pandas-dev__pandas-14886

5f889a2106f6584583458e01dbd0f3b9b696fab2

diff --git a/doc/source/whatsnew/v0.19.2.txt b/doc/source/whatsnew/v0.19.2.txt --- a/doc/source/whatsnew/v0.19.2.txt +++ b/doc/source/whatsnew/v0.19.2.txt @@ -78,7 +78,7 @@ Bug Fixes - Bug in clipboard functions on linux with python2 with unicode and separators (:issue:`13747`) - Bug in clipboard functions on Windows 10 and python 3 (:issue:`14362`, :issue:`12807`) - Bug in ``.to_clipboard()`` and Excel compat (:issue:`12529`) - +- Bug in ``DataFrame.combine_first()`` for integer columns (:issue:`14687`). - Bug in ``pd.read_csv()`` in which the ``dtype`` parameter was not being respected for empty data (:issue:`14712`) - Bug in ``pd.read_csv()`` in which the ``nrows`` parameter was not being respected for large input when using the C engine for parsing (:issue:`7626`) diff --git a/pandas/core/frame.py b/pandas/core/frame.py --- a/pandas/core/frame.py +++ b/pandas/core/frame.py @@ -3665,10 +3665,8 @@ def combine(self, other, func, fill_value=None, overwrite=True): otherSeries[other_mask] = fill_value # if we have different dtypes, possibily promote - if notnull(series).all(): - new_dtype = this_dtype - otherSeries = otherSeries.astype(new_dtype) - else: + new_dtype = this_dtype + if not is_dtype_equal(this_dtype, other_dtype): new_dtype = _find_common_type([this_dtype, other_dtype]) if not is_dtype_equal(this_dtype, new_dtype): series = series.astype(new_dtype)

combine_first throws ValueError: Cannot convert NA to integer I do not understand why there is a need to convert NA to integer if the result does not have NAs. Perhaps the combine_first algo needs to do it under the hood? #### A small, complete example of the issue ```python from pandas import DataFrame DataFrame({'a': [0, 1, 3, 5]}).combine_first(DataFrame({'a': [1, 4]})) Traceback (most recent call last): File "/usr/local/lib/python3.4/dist-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-16-12b973b1b150>", line 1, in <module> pd.DataFrame({'a': [0, 1, 3, 5]}).combine_first(pd.DataFrame({'a': [1, 4]})) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3787, in combine_first return self.combine(other, combiner, overwrite=False) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3714, in combine otherSeries = otherSeries.astype(new_dtype) File "/usr/local/lib/python3.4/dist-packages/pandas/core/generic.py", line 3054, in astype raise_on_error=raise_on_error, **kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3168, in astype return self.apply('astype', dtype=dtype, **kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3035, in apply applied = getattr(b, f)(**kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 462, in astype values=values, **kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 505, in _astype values = _astype_nansafe(values.ravel(), dtype, copy=True) File "/usr/local/lib/python3.4/dist-packages/pandas/types/cast.py", line 531, in _astype_nansafe raise ValueError('Cannot convert NA to integer') ValueError: Cannot convert NA to integer ``` #### Expected Output ``` a 0 0 1 1 2 3 3 5 ``` It does work when at least one item is a float: ```python DataFrame({'a': [0.0, 1, 3, 5]}).combine_first(DataFrame({'a': [1, 4]})) a 0 0.0 1 1.0 2 3.0 3 5.0 ``` I am aware that integer series cannot have NAs but there is no need to introduce NAs here. I do like it that the series is not upcasted to float silently though. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.4.3.final.0 python-bits: 64 OS: Linux OS-release: 3.19.0-66-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.19.0 nose: None pip: 1.5.4 setuptools: 3.3 Cython: 0.24.1 numpy: 1.11.2 scipy: 0.17.1 statsmodels: 0.6.1 xarray: None IPython: 4.0.0 sphinx: None patsy: 0.4.1 dateutil: 2.5.3 pytz: 2016.7 blosc: None bottleneck: None tables: 3.2.2 numexpr: 2.4.6 matplotlib: 1.5.1 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: 0.999 httplib2: 0.9.2 apiclient: None sqlalchemy: None pymysql: None psycopg2: 2.6.1 (dt dec pq3 ext lo64) jinja2: 2.8 boto: None pandas_datareader: None </details>

This seems to be a regression from 0.18, as this worked before: ``` In [1]: DataFrame({'a': [0, 1, 3, 5]}).combine_first(DataFrame({'a': [1, 4]})) Out[1]: a 0 0 1 1 2 3 3 5 In [2]: pd.__version__ Out[2]: u'0.18.1' ``` @Dmitrii-I Thanks for the report! Always welcome to look into what could have caused this change.

2016-12-15T10:53:24Z

[]

Traceback (most recent call last): File "/usr/local/lib/python3.4/dist-packages/IPython/core/interactiveshell.py", line 3066, in run_code exec(code_obj, self.user_global_ns, self.user_ns) File "<ipython-input-16-12b973b1b150>", line 1, in <module> pd.DataFrame({'a': [0, 1, 3, 5]}).combine_first(pd.DataFrame({'a': [1, 4]})) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3787, in combine_first return self.combine(other, combiner, overwrite=False) File "/usr/local/lib/python3.4/dist-packages/pandas/core/frame.py", line 3714, in combine otherSeries = otherSeries.astype(new_dtype) File "/usr/local/lib/python3.4/dist-packages/pandas/core/generic.py", line 3054, in astype raise_on_error=raise_on_error, **kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3168, in astype return self.apply('astype', dtype=dtype, **kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 3035, in apply applied = getattr(b, f)(**kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 462, in astype values=values, **kwargs) File "/usr/local/lib/python3.4/dist-packages/pandas/core/internals.py", line 505, in _astype values = _astype_nansafe(values.ravel(), dtype, copy=True) File "/usr/local/lib/python3.4/dist-packages/pandas/types/cast.py", line 531, in _astype_nansafe raise ValueError('Cannot convert NA to integer') ValueError: Cannot convert NA to integer

11,070

pandas-dev/pandas

pandas-dev__pandas-14907

e503d40ace473556990e5453ed5b4c9aa96e24ff

groupby/transform with NaNs in grouped column What's the expected behavior when grouping on a column containing `NaN` and then applying `transform`? For a `Series`, the current result is to throw an exception: ``` >>> df = pd.DataFrame({ ... 'a' : range(10), ... 'b' : [1, 1, 2, 3, np.nan, 4, 4, 5, 5, 5]}) >>> >>> df.groupby(df.b)['a'].transform(max) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/groupby.py", line 2422, in transform return self._transform_fast(cyfunc) File "pandas/core/groupby.py", line 2463, in _transform_fast return self._set_result_index_ordered(Series(values)) File "pandas/core/groupby.py", line 498, in _set_result_index_ordered result.index = self.obj.index File "pandas/core/generic.py", line 1997, in __setattr__ return object.__setattr__(self, name, value) File "pandas/src/properties.pyx", line 65, in pandas.lib.AxisProperty.__set__ (pandas/lib.c:41301) obj._set_axis(self.axis, value) File "pandas/core/series.py", line 273, in _set_axis self._data.set_axis(axis, labels) File "pandas/core/internals.py", line 2219, in set_axis 'new values have %d elements' % (old_len, new_len)) ValueError: Length mismatch: Expected axis has 9 elements, new values have 10 elements ``` For a `DataFrame`, the missing value gets filled in with what looks like an uninitialized value from `np.empty_like`: ``` >>> df.groupby(df.b).transform(max) a 0 1 1 1 2 2 3 3 4 -1 5 6 6 6 7 9 8 9 9 9 ``` It seems like either it should fill in the missing values with `NaN` (which might require a change of dtype), or just drop those rows from the result (which requires the shape to change). Either solution has the potential to surprise.

http://pandas.pydata.org/pandas-docs/stable/groupby.html#na-group-handling This _should_ work, so this is a bug as the NA group is not defined. Resultant value should be `NaN`. xref #5456 xref #6992 xref #443

2016-12-18T07:12:07Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/groupby.py", line 2422, in transform return self._transform_fast(cyfunc) File "pandas/core/groupby.py", line 2463, in _transform_fast return self._set_result_index_ordered(Series(values)) File "pandas/core/groupby.py", line 498, in _set_result_index_ordered result.index = self.obj.index File "pandas/core/generic.py", line 1997, in __setattr__ return object.__setattr__(self, name, value) File "pandas/src/properties.pyx", line 65, in pandas.lib.AxisProperty.__set__ (pandas/lib.c:41301) obj._set_axis(self.axis, value) File "pandas/core/series.py", line 273, in _set_axis self._data.set_axis(axis, labels) File "pandas/core/internals.py", line 2219, in set_axis 'new values have %d elements' % (old_len, new_len)) ValueError: Length mismatch: Expected axis has 9 elements, new values have 10 elements

11,072

pandas-dev/pandas

pandas-dev__pandas-14952

74de478392e09cf938d244f5990da4e001afc84c

Groupby.groups doesn't work by a groups convert from DateTimeIndex ``` i = pd.DatetimeIndex(pd.date_range('2015/01/01', periods=5), name='date') d = pd.DataFrame({'A':[5,6,7,8,9], 'B':[1,2,3,4,5]}, index=i) print i print d dg = d.groupby(level='date') print dg.get_group('2015-01-01') print dg.groups ``` **Output:** ``` DatetimeIndex(['2015-01-01', '2015-01-02', '2015-01-03', '2015-01-04', '2015-01-05'], dtype='datetime64[ns]', name=u'date', freq='D', tz=None) A B date 2015-01-01 5 1 2015-01-02 6 2 2015-01-03 7 3 2015-01-04 8 4 2015-01-05 9 5 A B date 2015-01-01 5 1 Traceback (most recent call last): File "<ipython-input-2-f2b4d1146750>", line 12, in <module> print dg.groups File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 405, in groups return self.grouper.groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 1352, in groups return self.groupings[0].groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 2041, in groups return self.index.groupby(self.grouper) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/tseries/base.py", line 60, in groupby return _algos.groupby_object(objs, f) TypeError: Argument 'labels' has incorrect type (expected numpy.ndarray, got DatetimeIndex) ```

hmm, does seem a bit buggy @Temmplar reason you closed this? @jreback sorry I guess, I'm not familiar with the system and I thought if you answered I have to close it. nope would be closed by a pull request to fix which are welcome! I tried this in `tseries/base.py`: ``` - return _algos.groupby_object(objs, f) + return _algos.groupby_object(objs, np.asarray(f)) ``` And it sort of works, but the results are ugly, especially if a timezone is involved. An index like this: ``` DatetimeIndex(['2016-06-28 05:30:00-05:00', '2016-06-28 05:31:00-05:00'], dtype='datetime64[ns, America/Chicago]') ``` Produces naive UTC results: ``` {numpy.datetime64('2016-06-28T10:30:00.000000000'): [Timestamp('2016-06-28 05:30:00-0500', ... ``` @jreback Do you have any idea how to fix this simply and properly? I am getting this error too.

2016-12-22T06:45:33Z

[]

Traceback (most recent call last): File "<ipython-input-2-f2b4d1146750>", line 12, in <module> print dg.groups File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 405, in groups return self.grouper.groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 1352, in groups return self.groupings[0].groups File "pandas/src/properties.pyx", line 34, in pandas.lib.cache_readonly.__get__ (pandas/lib.c:41917) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/core/groupby.py", line 2041, in groups return self.index.groupby(self.grouper) File "/home/george/anaconda/lib/python2.7/site-packages/pandas/tseries/base.py", line 60, in groupby return _algos.groupby_object(objs, f) TypeError: Argument 'labels' has incorrect type (expected numpy.ndarray, got DatetimeIndex)

11,079

pandas-dev/pandas

pandas-dev__pandas-15569

5f0b69aee3622eed9392cef163e4b31ba742498e

diff --git a/doc/source/whatsnew/v0.20.0.txt b/doc/source/whatsnew/v0.20.0.txt --- a/doc/source/whatsnew/v0.20.0.txt +++ b/doc/source/whatsnew/v0.20.0.txt @@ -635,7 +635,7 @@ Performance Improvements - Increased performance of ``pd.factorize()`` by releasing the GIL with ``object`` dtype when inferred as strings (:issue:`14859`) - Improved performance of timeseries plotting with an irregular DatetimeIndex (or with ``compat_x=True``) (:issue:`15073`). -- Improved performance of ``groupby().cummin()`` and ``groupby().cummax()`` (:issue:`15048`, :issue:`15109`) +- Improved performance of ``groupby().cummin()`` and ``groupby().cummax()`` (:issue:`15048`, :issue:`15109`, :issue:`15561`) - Improved performance and reduced memory when indexing with a ``MultiIndex`` (:issue:`15245`) - When reading buffer object in ``read_sas()`` method without specified format, filepath string is inferred rather than buffer object. (:issue:`14947`) - Improved performance of `rank()` for categorical data (:issue:`15498`) diff --git a/pandas/core/groupby.py b/pandas/core/groupby.py --- a/pandas/core/groupby.py +++ b/pandas/core/groupby.py @@ -1442,7 +1442,7 @@ def cummin(self, axis=0, **kwargs): if axis != 0: return self.apply(lambda x: np.minimum.accumulate(x, axis)) - return self._cython_transform('cummin', **kwargs) + return self._cython_transform('cummin', numeric_only=False) @Substitution(name='groupby') @Appender(_doc_template) @@ -1451,7 +1451,7 @@ def cummax(self, axis=0, **kwargs): if axis != 0: return self.apply(lambda x: np.maximum.accumulate(x, axis)) - return self._cython_transform('cummax', **kwargs) + return self._cython_transform('cummax', numeric_only=False) @Substitution(name='groupby') @Appender(_doc_template)

BUG: cython version of groupby.cummax throws error on datetimes #### Code Sample, a copy-pastable example if possible ```python >>> import pandas as pd >>> x = pd.DataFrame(dict(a=[1], b=pd.to_datetime(['2001']))) >>> x.groupby('a').b.cummax() Traceback (most recent call last): File "<ipython-input-9-316257648d5f>", line 1, in <module> x.groupby('a').b.cummax() File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1454, in cummax return self._cython_transform('cummax', **kwargs) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 806, in _cython_transform raise DataError('No numeric types to aggregate') DataError: No numeric types to aggregate ``` #### Problem description The current github version of pandas has cython implementations of `groupby.cummin` and `groupby.cummax`, which throw an error if called on datetime columns. (See #15048, 0fe491d.) #### Expected Output ```python 0 2001-01-01 Name: b, dtype: datetime64[ns] ``` #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.5.2.final.0 python-bits: 64 OS: Linux OS-release: 4.9.8-100.fc24.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: C LANG: C LOCALE: None.None pandas: 0.19.0+531.g04e1168 pytest: 3.0.5 pip: 9.0.1 setuptools: 27.2.0 Cython: 0.25.2 numpy: 1.11.3 scipy: 0.18.1 xarray: 0.9.1 IPython: 4.2.0 sphinx: 1.5.1 patsy: 0.4.1 dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.2.0 tables: 3.3.0 numexpr: 2.6.2 feather: None matplotlib: 2.0.0 openpyxl: 2.4.1 xlrd: 1.0.0 xlwt: 1.2.0 xlsxwriter: 0.9.6 lxml: 3.7.2 bs4: 4.5.3 html5lib: 0.999 sqlalchemy: 1.1.5 pymysql: None psycopg2: None jinja2: 2.9.4 s3fs: None pandas_gbq: None pandas_datareader: None </details>

I think this may be as easy as adding a `numeric_only=False` to this function call. (of course with tests) https://github.com/pandas-dev/pandas/blob/04e116851337cd852b4255f8221d9be44829e0e1/pandas/core/groupby.py#L1454 This is correct and not a bug. datetimes are not numeric types. This has to be specifically enabled. this was covered by #15054 ``` In [29]: x = pd.DataFrame(dict(a=[1], b=pd.to_datetime(['2001']))) ...: >>> x.groupby('a').b.cummax(numeric_only=False) ...: ...: Out[29]: 0 2001-01-01 Name: b, dtype: datetime64[ns] ``` @jreback - maybe the API is set, but isn't this inconsistent? ```python In [23]: df = pd.DataFrame(dict(a=[1], b=pd.to_datetime(['2001']))) In [24]: df['b'].max() Out[24]: Timestamp('2001-01-01 00:00:00') In [25]: df['b'].cummax() Out[25]: 0 2001-01-01 Name: b, dtype: datetime64[ns] In [26]: df.groupby('a')['b'].max() Out[26]: a 1 2001-01-01 Name: b, dtype: datetime64[ns] In [27]: df.groupby('a')['b'].cummax() Out[27]: DataError: No numeric types to aggregate ``` @chris-b1 hmm, ahh we are defaulting ``numeric_only=False`` for ``.max`` yes that should be done for ``cummax``/``cummin``.

2017-03-04T20:20:15Z

[]

Traceback (most recent call last): File "<ipython-input-9-316257648d5f>", line 1, in <module> x.groupby('a').b.cummax() File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1454, in cummax return self._cython_transform('cummax', **kwargs) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 806, in _cython_transform raise DataError('No numeric types to aggregate') DataError: No numeric types to aggregate

11,116

pandas-dev/pandas

pandas-dev__pandas-16090

d313e4dd7605a658869f5d026d6705afb169ab40

diff --git a/doc/source/whatsnew/v0.20.0.txt b/doc/source/whatsnew/v0.20.0.txt --- a/doc/source/whatsnew/v0.20.0.txt +++ b/doc/source/whatsnew/v0.20.0.txt @@ -1600,7 +1600,7 @@ Indexing - Bug in the HTML display with with a ``MultiIndex`` and truncation (:issue:`14882`) - Bug in the display of ``.info()`` where a qualifier (+) would always be displayed with a ``MultiIndex`` that contains only non-strings (:issue:`15245`) - Bug in ``pd.concat()`` where the names of ``MultiIndex`` of resulting ``DataFrame`` are not handled correctly when ``None`` is presented in the names of ``MultiIndex`` of input ``DataFrame`` (:issue:`15787`) -- Bug in ``DataFrame.sort_index()`` and ``Series.sort_index()`` where ``na_position`` doesn't work with a ``MultiIndex`` (:issue:`14784`) +- Bug in ``DataFrame.sort_index()`` and ``Series.sort_index()`` where ``na_position`` doesn't work with a ``MultiIndex`` (:issue:`14784`, :issue:`16604`) I/O ^^^ diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py --- a/pandas/core/indexes/multi.py +++ b/pandas/core/indexes/multi.py @@ -1645,10 +1645,11 @@ def _get_labels_for_sorting(self): """ from pandas.core.categorical import Categorical - return [Categorical.from_codes(label, - np.arange(np.array(label).max() + 1, - dtype=label.dtype), - ordered=True) + def cats(label): + return np.arange(np.array(label).max() + 1 if len(label) else 0, + dtype=label.dtype) + + return [Categorical.from_codes(label, cats(label), ordered=True) for label in self.labels] def sortlevel(self, level=0, ascending=True, sort_remaining=True):

BUG: groupy().nth() throws error on multiple groups, empty result #### Code Sample, a copy-pastable example if possible ```python >>> import pandas as pd >>> df = pd.DataFrame(index=[0], columns=['a', 'b', 'c']) >>> df.groupby('a').nth(10) Empty DataFrame Columns: [b, c] Index: [] >>> df.groupby(['a', 'b']).nth(10) Traceback (most recent call last): File "<ipython-input-3-ae8299c3984e>", line 1, in <module> df.groupby(['a', 'b']).nth(10) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1390, in nth return out.sort_index() if self.sort else out File "~/anaconda3/lib/python3.5/site-packages/pandas/core/frame.py", line 3344, in sort_index indexer = lexsort_indexer(labels._get_labels_for_sorting(), File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in _get_labels_for_sorting for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in <listcomp> for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/numpy/core/_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) ValueError: zero-size array to reduction operation maximum which has no identity ``` #### Problem description In the current Github version of Pandas, when calling `groupby().nth()` with multiple grouping columns, an error is raised if the result is empty. This is a regression from version 0.19.2. #### Expected Output ```python Empty DataFrame Columns: [b, c] Index: [] Empty DataFrame Columns: [c] Index: [] ``` #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.5.2.final.0 python-bits: 64 OS: Linux OS-release: 4.9.8-100.fc24.x86_64 machine: x86_64 processor: x86_64 byteorder: little LC_ALL: C LANG: C LOCALE: None.None pandas: 0.19.0+829.gb17e286 pytest: 3.0.5 pip: 9.0.1 setuptools: 27.2.0 Cython: 0.25.2 numpy: 1.11.3 scipy: 0.18.1 xarray: 0.9.1 IPython: 4.2.0 sphinx: 1.5.1 patsy: 0.4.1 dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.2.0 tables: 3.3.0 numexpr: 2.6.2 feather: None matplotlib: 2.0.0 openpyxl: 2.4.1 xlrd: 1.0.0 xlwt: 1.2.0 xlsxwriter: 0.9.6 lxml: 3.7.2 bs4: 4.5.3 html5lib: 0.999 sqlalchemy: 1.1.5 pymysql: None psycopg2: None jinja2: 2.9.4 s3fs: None pandas_gbq: None pandas_datareader: None </details>

Think this just requires: ```diff diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py index 92baf9d..34b62c5 100644 --- a/pandas/core/indexes/multi.py +++ b/pandas/core/indexes/multi.py @@ -1645,11 +1645,9 @@ class MultiIndex(Index): """ from pandas.core.categorical import Categorical - return [Categorical.from_codes(label, - np.arange(np.array(label).max() + 1, - dtype=label.dtype), - ordered=True) - for label in self.labels] + return [Categorical.from_codes(label, np.arange( + np.array(label).max() + 1 if len(label) else 0, + dtype=label.dtype), ordered=True) for label in self.labels] def sortlevel(self, level=0, ascending=True, sort_remaining=True): """ ``` can u put up a PR with that fix (and test)?

2017-04-21T23:11:51Z

[]

Traceback (most recent call last): File "<ipython-input-3-ae8299c3984e>", line 1, in <module> df.groupby(['a', 'b']).nth(10) File "~/anaconda3/lib/python3.5/site-packages/pandas/core/groupby.py", line 1390, in nth return out.sort_index() if self.sort else out File "~/anaconda3/lib/python3.5/site-packages/pandas/core/frame.py", line 3344, in sort_index indexer = lexsort_indexer(labels._get_labels_for_sorting(), File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in _get_labels_for_sorting for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/pandas/core/indexes/multi.py", line 1652, in <listcomp> for label in self.labels] File "~/anaconda3/lib/python3.5/site-packages/numpy/core/_methods.py", line 26, in _amax return umr_maximum(a, axis, None, out, keepdims) ValueError: zero-size array to reduction operation maximum which has no identity

11,152

pandas-dev/pandas

pandas-dev__pandas-16294

4bed864a24901d9c2baab5e17c57c956a188602f

diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -48,7 +48,7 @@ Indexing I/O ^^^ - +- Bug that would force importing of the clipboard routines unecessarily, potentially causing an import error on startup (:issue:`16288`) Plotting diff --git a/pandas/core/generic.py b/pandas/core/generic.py --- a/pandas/core/generic.py +++ b/pandas/core/generic.py @@ -1382,8 +1382,8 @@ def to_clipboard(self, excel=None, sep=None, **kwargs): - Windows: none - OS X: none """ - from pandas.io.clipboard import clipboard - clipboard.to_clipboard(self, excel=excel, sep=sep, **kwargs) + from pandas.io import clipboards + clipboards.to_clipboard(self, excel=excel, sep=sep, **kwargs) def to_xarray(self): """ diff --git a/pandas/io/api.py b/pandas/io/api.py --- a/pandas/io/api.py +++ b/pandas/io/api.py @@ -5,7 +5,7 @@ # flake8: noqa from pandas.io.parsers import read_csv, read_table, read_fwf -from pandas.io.clipboard.clipboard import read_clipboard +from pandas.io.clipboards import read_clipboard from pandas.io.excel import ExcelFile, ExcelWriter, read_excel from pandas.io.pytables import HDFStore, get_store, read_hdf from pandas.io.json import read_json diff --git a/pandas/io/clipboard/clipboard.py b/pandas/io/clipboards.py similarity index 100% rename from pandas/io/clipboard/clipboard.py rename to pandas/io/clipboards.py

ImportError with pandas 0.20.0 and 0.20.1 #### Code Sample, a copy-pastable example if possible With versions 0.20.0 and 0.20.1, I get the following error: ```python import pandas ``` ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/__init__.py", line 58, in <module> from pandas.io.api import * File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/api.py", line 8, in <module> from pandas.io.clipboard.clipboard import read_clipboard File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 103, in <module> copy, paste = determine_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 76, in determine_clipboard return init_qt_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/clipboards.py", line 49, in init_qt_clipboard from PyQt4.QtGui import QApplication ImportError: No module named sip ``` This does not occur with 0.19.2 Here are the other packages installed in my virtualenv ``` $ pip freeze appdirs==1.4.3 numpy==1.12.1 packaging==16.8 pandas==0.20.1 pyparsing==2.2.0 python-dateutil==2.6.0 pytz==2017.2 PyYAML==3.12 six==1.10.0 ``` #### Problem description It would be nice if pandas was pip installable. As it stands, the new versions are not (for me). #### Expected Output No error when importing pandas. #### Output of ``pd.show_versions()`` I can't get that info without being able to import pandas.

@jorisvandenbossche did you see something like this earlier? Not exactly this error, but possibly related. For me the error with clipboard was something with ``from PyQt4 import QtCore``. For some reason, I had an empty PyQt4 package in site-packages next to PyQt5 (so `import PyQt4` did work, which is used to check which clipboard backend to use, but then later on actual imports raise an error), and for some reason this started to give problems. Matplotlib also had this problem, so I am not sure it was pandas-related. @FragLegs Is this in a clean (newly made) env? Or if not, can you see if you can reproduce it then as well?

2017-05-09T01:07:17Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/__init__.py", line 58, in <module> from pandas.io.api import * File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/api.py", line 8, in <module> from pandas.io.clipboard.clipboard import read_clipboard File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 103, in <module> copy, paste = determine_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/__init__.py", line 76, in determine_clipboard return init_qt_clipboard() File "/home/smiel/.venvs/foo/local/lib/python2.7/site-packages/pandas/io/clipboard/clipboards.py", line 49, in init_qt_clipboard from PyQt4.QtGui import QApplication ImportError: No module named sip

11,183

pandas-dev/pandas

pandas-dev__pandas-16434

49ec31bbaeca81a6f58fc1be26fe80f3ac188cdd

diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -80,7 +80,7 @@ Reshaping ^^^^^^^^^ - Bug in ``DataFrame.stack`` with unsorted levels in MultiIndex columns (:issue:`16323`) - +- Bug in ``Series.isin(..)`` with a list of tuples (:issue:`16394`) Numeric ^^^^^^^ diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -388,7 +388,7 @@ def isin(comps, values): "[{0}]".format(type(values).__name__)) if not isinstance(values, (ABCIndex, ABCSeries, np.ndarray)): - values = np.array(list(values), dtype='object') + values = lib.list_to_object_array(list(values)) comps, dtype, _ = _ensure_data(comps) values, _, _ = _ensure_data(values, dtype=dtype)

BUG: ValueError with Series.isin and tuples #### Code Sample, a copy-pastable example if possible ``` import pandas as pd df = pd.DataFrame({'A': [1, 2, 3], 'B': ['a', 'b', 'f']}) df['C'] = list(zip(df['A'], df['B'])) df['C'].isin([(1, 'a')]) ``` #### Problem description Returns ValueError: Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/series.py", line 2555, in isin result = algorithms.isin(_values_from_object(self), values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 421, in isin return f(comps, values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 399, in <lambda> f = lambda x, y: htable.ismember_object(x, values) File "pandas/_libs/hashtable_func_helper.pxi", line 428, in pandas._libs.hashtable.ismember_object (pandas/_libs/hashtable.c:29677) ValueError: Buffer has wrong number of dimensions (expected 1, got 2) #### Expected Output In pandas 0.19.2 returns: 0 True 1 False 2 False Name: C, dtype: bool #### Output of ``pd.show_versions()`` <details> # Paste the output here pd.show_versions() here INSTALLED VERSIONS ------------------ commit: None python: 3.6.1.final.0 python-bits: 64 OS: Darwin OS-release: 16.5.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.20.0rc2 pytest: None pip: 9.0.1 setuptools: 27.2.0 Cython: None numpy: 1.12.1 scipy: None xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None </details>

this code was refactored to be more general, so this was a missing case. easy fix I think. ``np.array`` converts nested tuples to lists, which is not nice, so do this. if you'd like to submit a PR with this as an added tests (and make sure nothing else breaks), would be great. ``` diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py index a745ec6..77d79c9 100644 --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -388,7 +388,7 @@ def isin(comps, values): "[{0}]".format(type(values).__name__)) if not isinstance(values, (ABCIndex, ABCSeries, np.ndarray)): - values = np.array(list(values), dtype='object') + values = lib.list_to_object_array(list(values)) comps, dtype, _ = _ensure_data(comps) values, _, _ = _ensure_data(values, dtype=dtype) ``` I'm taking a crack at this. Is the solution to just add lib.list_to_object_array back in along with a test for the tuple case, or should we check if comps contains tuples and use lib.list_to_object_array only if it does?

2017-05-22T20:03:21Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/series.py", line 2555, in isin result = algorithms.isin(_values_from_object(self), values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 421, in isin return f(comps, values) File "/anaconda/envs/pandas_dev/lib/python3.6/site-packages/pandas/core/algorithms.py", line 399, in <lambda> f = lambda x, y: htable.ismember_object(x, values) File "pandas/_libs/hashtable_func_helper.pxi", line 428, in pandas._libs.hashtable.ismember_object (pandas/_libs/hashtable.c:29677) ValueError: Buffer has wrong number of dimensions (expected 1, got 2)

11,206

pandas-dev/pandas

pandas-dev__pandas-16486

92d07992e826808cd56f0bd8fec083b510ca402d

diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -40,6 +40,7 @@ Bug Fixes - Silenced a warning on some Windows environments about "tput: terminal attributes: No such device or address" when detecting the terminal size. This fix only applies to python 3 (:issue:`16496`) - Bug in using ``pathlib.Path`` or ``py.path.local`` objects with io functions (:issue:`16291`) +- Bug in ``Index.symmetric_difference()`` on two equal MultiIndex's, results in a TypeError (:issue `13490`) - Bug in ``DataFrame.update()`` with ``overwrite=False`` and ``NaN values`` (:issue:`15593`) diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py --- a/pandas/core/indexes/multi.py +++ b/pandas/core/indexes/multi.py @@ -414,6 +414,12 @@ def view(self, cls=None): return result def _shallow_copy_with_infer(self, values=None, **kwargs): + # On equal MultiIndexes the difference is empty. + # Therefore, an empty MultiIndex is returned GH13490 + if len(values) == 0: + return MultiIndex(levels=[[] for _ in range(self.nlevels)], + labels=[[] for _ in range(self.nlevels)], + **kwargs) return self._shallow_copy(values, **kwargs) @Appender(_index_shared_docs['_shallow_copy'])

Symmetric difference on equal MultiIndexes raises TypeError Calling `symmetric_difference` on two equal multiindices results in a TypeError rather than an empty MultiIndex. This is surprising since calling `difference` on the same multiindices results in the expected empty MultiIndex. #### Code Sample, a copy-pastable example if possible ``` a = pandas.MultiIndex.from_product([['a', 'b'], [0, 1]]) b = pandas.MultiIndex.from_product([['a', 'b'], [0, 1]]) print(a.symmetric_difference(b)) ``` Which gives the following error: ``` Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1674, in __xor__ return self.symmetric_difference(other) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1911, in symmetric_difference return self._shallow_copy_with_infer(the_diff, **attribs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 387, in _shallow_copy_with_infer return self._shallow_copy(values, **kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 396, in _shallow_copy return MultiIndex.from_tuples(values, **kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 883, in from_tuples raise TypeError('Cannot infer number of levels from empty list') TypeError: Cannot infer number of levels from empty list ``` #### Expected Output ``` MultiIndex(levels=[[], []], labels = [[], []]) ``` #### output of `pd.show_versions()` ``` INSTALLED VERSIONS ------------------ commit: None python: 2.7.11.final.0 python-bits: 64 OS: Windows OS-release: 7 machine: AMD64 processor: Intel64 Family 6 Model 61 Stepping 4, GenuineIntel byteorder: little LC_ALL: None LANG: DA pandas: 0.18.1 nose: None pip: 8.1.1 setuptools: 20.7.0 Cython: 0.23 numpy: 1.10.4 scipy: 0.17.0 statsmodels: 0.6.1 xarray: None IPython: 4.2.0 sphinx: None patsy: 0.4.1 dateutil: 2.5.2 pytz: 2016.3 blosc: None bottleneck: None tables: 3.2.2 numexpr: 2.5.2 matplotlib: 1.5.1 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None httplib2: None apiclient: None sqlalchemy: 1.0.12 pymysql: None psycopg2: None jinja2: 2.8 boto: None pandas_datareader: None ```

yeah I agree, on empties this should behave the same as `.difference` / `.union` cc @TomAugspurger Agreed.

2017-05-24T20:40:41Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1674, in __xor__ return self.symmetric_difference(other) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\base.py", line 1911, in symmetric_difference return self._shallow_copy_with_infer(the_diff, **attribs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 387, in _shallow_copy_with_infer return self._shallow_copy(values, **kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 396, in _shallow_copy return MultiIndex.from_tuples(values, **kwargs) File "C:\Miniconda3\envs\balancedbeta\lib\site-packages\pandas\indexes\multi.py", line 883, in from_tuples raise TypeError('Cannot infer number of levels from empty list') TypeError: Cannot infer number of levels from empty list

11,222

pandas-dev/pandas

pandas-dev__pandas-16526

ef487d9e474e8052c0f7c6260de5802a950defad

Various py3k test failures in tests.io.test_html with US-ASCII preferred encoding #### Code Sample, a copy-pastable example if possible ```python >>> import locale >>> locale.getpreferredencoding() 'US-ASCII' >>> open('/usr/local/lib/python3.4/site-packages/pandas/tests/io/data/spam.html').read() Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python3.4/encodings/ascii.py", line 26, in decode return codecs.ascii_decode(input, self.errors)[0] UnicodeDecodeError: 'ascii' codec can't decode byte 0xc2 in position 21552: ordinal not in range(128) >>> ``` #### Problem description Three tests, `test_string_io`, `test_string`, and `test_file_like`, all open `spam.html` without specifying the encoding, and then attempt to read it. This causes the tests to terminate prematurely with an error. #### Expected Output All three tests should pass since the code under test is not responsible for determining the file encoding. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.4.6.final.0 python-bits: 64 OS: FreeBSD OS-release: 10.3-STABLE machine: amd64 processor: amd64 byteorder: little LC_ALL: None LANG: None LOCALE: None.None pandas: 0.20.1 pytest: 3.1.0 pip: None setuptools: 32.1.0 Cython: None numpy: 1.11.2 scipy: 0.19.0 xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.0.0 tables: 3.4.2 numexpr: 2.6.2 feather: None matplotlib: None openpyxl: 2.4.7 xlrd: 1.0.0 xlwt: None xlsxwriter: 0.9.6 lxml: 3.6.0 bs4: 4.5.1 html5lib: 0.9999999 sqlalchemy: 1.1.10 pymysql: 0.7.11.None psycopg2: 2.7.1 (dt dec pq3 ext lo64) jinja2: 2.9.5 s3fs: None pandas_gbq: None pandas_datareader: None </details>

2017-05-28T18:59:57Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python3.4/encodings/ascii.py", line 26, in decode return codecs.ascii_decode(input, self.errors)[0] UnicodeDecodeError: 'ascii' codec can't decode byte 0xc2 in position 21552: ordinal not in range(128)

11,231

pandas-dev/pandas

pandas-dev__pandas-16543

03d44f3dd0ffd55d7538b67466cf4d3899ceac27

diff --git a/doc/source/whatsnew/v0.20.2.txt b/doc/source/whatsnew/v0.20.2.txt --- a/doc/source/whatsnew/v0.20.2.txt +++ b/doc/source/whatsnew/v0.20.2.txt @@ -44,8 +44,7 @@ Bug Fixes - Bug in ``DataFrame.update()`` with ``overwrite=False`` and ``NaN values`` (:issue:`15593`) - Passing an invalid engine to :func:`read_csv` now raises an informative ``ValueError`` rather than ``UnboundLocalError``. (:issue:`16511`) - - +- Bug in :func:`unique` on an array of tuples (:issue:`16519`) - Fixed a compatibility issue with IPython 6.0's tab completion showing deprecation warnings on Categoricals (:issue:`16409`) diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -163,7 +163,7 @@ def _ensure_arraylike(values): ABCIndexClass, ABCSeries)): inferred = lib.infer_dtype(values) if inferred in ['mixed', 'string', 'unicode']: - values = np.asarray(values, dtype=object) + values = lib.list_to_object_array(values) else: values = np.asarray(values) return values @@ -328,6 +328,11 @@ def unique(values): [b, a, c] Categories (3, object): [a < b < c] + An array of tuples + + >>> pd.unique([('a', 'b'), ('b', 'a'), ('a', 'c'), ('b', 'a')]) + array([('a', 'b'), ('b', 'a'), ('a', 'c')], dtype=object) + See Also -------- pandas.Index.unique

Regression from 0.19.2 to 0.20.1 in pandas.unique() when applied to list of tuples #### Code Sample, a copy-pastable example if possible ```python import pandas as pd input = [(0, 0), (0, 1), (1, 0), (1, 1), (0, 0), (0, 1), (1, 0), (1, 1)] print pd.unique(input) ``` #### Problem description The code exits unexpectedly ``` Traceback (most recent call last): File "pandas_bug.py", line 6, in <module> pd.unique(input) File "/Users/johannes/.virtualenvs/pandas/lib/python2.7/site-packages/pandas/core/algorithms.py", line 351, in unique uniques = table.unique(values) File "pandas/_libs/hashtable_class_helper.pxi", line 1271, in pandas._libs.hashtable.PyObjectHashTable.unique (pandas/_libs/hashtable.c:21384) ValueError: Buffer has wrong number of dimensions (expected 1, got 2) ``` #### Expected Output The code works on pandas version 0.19.2 and produces the expected output ``` [(0, 0) (0, 1) (1, 0) (1, 1)] ``` Moreover this problem is not limited to MacOSX, but was also encounter on Ubuntu CI server. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Darwin OS-release: 15.6.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: None.None pandas: 0.20.1 pytest: None pip: 9.0.1 setuptools: 35.0.2 Cython: None numpy: 1.12.1 scipy: None xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None None </details>

this is related to #16394 and needs the same fix, along with some tests; ensuring that nothing else breaks. ``` diff --git a/pandas/core/algorithms.py b/pandas/core/algorithms.py index 77d79c9..9cfaf04 100644 --- a/pandas/core/algorithms.py +++ b/pandas/core/algorithms.py @@ -163,7 +163,7 @@ def _ensure_arraylike(values): ABCIndexClass, ABCSeries)): inferred = lib.infer_dtype(values) if inferred in ['mixed', 'string', 'unicode']: - values = np.asarray(values, dtype=object) + values = lib.list_to_object_array(values) else: values = np.asarray(values) return values ```

2017-05-30T22:05:37Z

[]

Traceback (most recent call last): File "pandas_bug.py", line 6, in <module> pd.unique(input) File "/Users/johannes/.virtualenvs/pandas/lib/python2.7/site-packages/pandas/core/algorithms.py", line 351, in unique uniques = table.unique(values) File "pandas/_libs/hashtable_class_helper.pxi", line 1271, in pandas._libs.hashtable.PyObjectHashTable.unique (pandas/_libs/hashtable.c:21384) ValueError: Buffer has wrong number of dimensions (expected 1, got 2)

11,236

pandas-dev/pandas

pandas-dev__pandas-16701

b7e7fd3f17d4d2a2f87b9d169cf87143f04e5d33

diff --git a/doc/source/whatsnew/v0.20.3.txt b/doc/source/whatsnew/v0.20.3.txt --- a/doc/source/whatsnew/v0.20.3.txt +++ b/doc/source/whatsnew/v0.20.3.txt @@ -59,7 +59,7 @@ I/O Plotting ^^^^^^^^ - +- Fix regression in series plotting that prevented RGB and RGBA tuples from being used as color arguments (:issue:`16233`) diff --git a/pandas/plotting/_core.py b/pandas/plotting/_core.py --- a/pandas/plotting/_core.py +++ b/pandas/plotting/_core.py @@ -187,6 +187,11 @@ def _validate_color_args(self): # support series.plot(color='green') self.kwds['color'] = [self.kwds['color']] + if ('color' in self.kwds and isinstance(self.kwds['color'], tuple) and + self.nseries == 1 and len(self.kwds['color']) in (3, 4)): + # support RGB and RGBA tuples in series plot + self.kwds['color'] = [self.kwds['color']] + if ('color' in self.kwds or 'colors' in self.kwds) and \ self.colormap is not None: warnings.warn("'color' and 'colormap' cannot be used "

DataFrame plot method no long takes RGB tuple as color arg #### Problem description The `plot` method on `DataFrame` objects takes a `color` argument that in versions prior to 0.20.2 took an RGB tuple as an accepted value. The 0.20.2 release throws an exception when specifying an RGB tuple for the `color` arg. #### Code Sample ```python # import matplotlib.pyplot as plt df = pandas.DataFrame([[1, 2], [3, 4]], columns=['a', 'b']) df.plot(x='b', y='a', color=(1, 0, 0)) plt.show() ``` #### Expected Output The expected output is a red line plot. The following exception is thrown: ```python Exception in Tkinter callback Traceback (most recent call last): File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/lib-tk/Tkinter.py", line 1536, in __call__ return self.func(*args) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 280, in resize self.show() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 351, in draw FigureCanvasAgg.draw(self) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 464, in draw self.figure.draw(self.renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, *args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/figure.py", line 1144, in draw renderer, self, dsu, self.suppressComposite) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, *args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/axes/_base.py", line 2426, in draw mimage._draw_list_compositing_images(renderer, self, dsu) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, *args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 803, in draw ln_color_rgba = self._get_rgba_ln_color() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 1344, in _get_rgba_ln_color return mcolors.to_rgba(self._color, self._alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 143, in to_rgba rgba = _to_rgba_no_colorcycle(c, alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 194, in _to_rgba_no_colorcycle raise ValueError("Invalid RGBA argument: {!r}".format(orig_c)) ValueError: Invalid RGBA argument: 1 ``` This plot is correctly displayed with an identical environment that has pandas 0.19.2 installed so is API breaking. Apologies if this is fixed already in master. #### Output of ``pd.show_versions()`` <details> NSTALLED VERSIONS ------------------ commit: None python: 2.7.10.final.0 python-bits: 64 OS: Darwin OS-release: 15.2.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.20.2 pytest: None pip: 9.0.1 setuptools: 36.0.1 Cython: None numpy: 1.13.0 scipy: None xarray: None IPython: 5.4.1 sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: 2.0.2 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None </details>

Follow up: Enclosing the tuple in a list works: ``` df.plot(x='b', y='a', color=[(1, 0, 0)]) ``` This looks related to #16233 which fixes #3486 but breaks what I perceive to be a fairly standard way of assigning line color (e.g. it is the first method listed for specifying color on https://matplotlib.org/users/colors.html). @kjford Thanks for the report! That is indeed a regression, and should be fixed. Want to do a PR to fix?

2017-06-15T06:33:34Z

[]

Traceback (most recent call last): File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/lib-tk/Tkinter.py", line 1536, in __call__ return self.func(*args) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 280, in resize self.show() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_tkagg.py", line 351, in draw FigureCanvasAgg.draw(self) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py", line 464, in draw self.figure.draw(self.renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, *args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/figure.py", line 1144, in draw renderer, self, dsu, self.suppressComposite) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, *args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/axes/_base.py", line 2426, in draw mimage._draw_list_compositing_images(renderer, self, dsu) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/image.py", line 139, in _draw_list_compositing_images a.draw(renderer) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/artist.py", line 63, in draw_wrapper draw(artist, renderer, *args, **kwargs) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 803, in draw ln_color_rgba = self._get_rgba_ln_color() File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/lines.py", line 1344, in _get_rgba_ln_color return mcolors.to_rgba(self._color, self._alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 143, in to_rgba rgba = _to_rgba_no_colorcycle(c, alpha) File "/venvs/pandas0.20.2/lib/python2.7/site-packages/matplotlib/colors.py", line 194, in _to_rgba_no_colorcycle raise ValueError("Invalid RGBA argument: {!r}".format(orig_c)) ValueError: Invalid RGBA argument: 1

11,258

pandas-dev/pandas

pandas-dev__pandas-16744

8a98f5ed541c87a9bf101c9331bd6cfa8f007cc9

Different behaviour on two different environments. TypeError: data type "datetime" not understood #### I have an aggregation logic ```python In [19]: datewise_prices Out[19]: [{'arrivalDate': 1490227200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490400000000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490832000000, 'maxPrice': 3500, 'minPrice': 3500, 'modalPrice': 3500}, {'arrivalDate': 1490918400000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1491091200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}] In [20]: weekly_dataframe = pandas.DataFrame(datewise_prices) ...: weekly_dataframe.drop('minPrice', axis=1, inplace=True) ...: weekly_dataframe.drop('maxPrice', axis=1, inplace=True) ...: weekly_dataframe['arrivalDate'] = pandas.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') ...: weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() ...: weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arriv ...: alDate'].dt.strftime('%d %b') ``` #### It results in an error on my Staging environment but works fine on my local. The staging environment gives an error in `(weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b')` ``` Traceback (most recent call last): File "/root/myapp/myapp/handlers/aggregation_handler.py", line 290, in get_daily_weekly_and_monthly_aggregates weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 696, in wrapper converted = _Op.get_op(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 330, in get_op return _TimeOp(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 343, in __init__ lvalues = self._convert_to_array(left, name=name) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 454, in _convert_to_array if (inferred_type in ('datetime64', 'datetime', 'date', 'time') or TypeError: data type "datetime" not understood ``` but it works fine on my local. #### Output of ``pd.show_versions()`` on local <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.6.final.0 python-bits: 64 OS: Linux OS-release: 4.4.0-78-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_IN LOCALE: None.None pandas: 0.19.2 nose: None pip: 9.0.1 setuptools: 36.0.1 Cython: None numpy: 1.12.1 scipy: None statsmodels: None xarray: None IPython: 5.4.1 sphinx: None patsy: None dateutil: 2.5.3 pytz: 2016.7 blosc: None bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: 0.9.6 lxml: None bs4: None html5lib: 0.999 httplib2: 0.8 apiclient: None sqlalchemy: None pymysql: None psycopg2: None jinja2: 2.8 boto: 2.40.0 pandas_datareader: None </details> #### Output of ``pd.show_versions()`` on Staging <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.6.final.0 python-bits: 64 OS: Linux OS-release: 3.13.0-57-generic machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: None.None pandas: 0.19.2 nose: None pip: 1.5.4 setuptools: 3.3 Cython: None numpy: 1.13.0 scipy: None statsmodels: None xarray: None IPython: 4.0.0 sphinx: None patsy: None dateutil: 2.5.3 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: 0.9.6 lxml: None bs4: None html5lib: 0.999 httplib2: None apiclient: None sqlalchemy: 1.0.8 pymysql: None psycopg2: None jinja2: 2.9.6 boto: 2.40.0 pandas_datareader: None </details> Please help me out.

Can you make a copy-pastable example, as requested? ``` import pandas as pd datewise_prices = [ {'arrivalDate': 1490227200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490400000000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490832000000, 'maxPrice': 3500, 'minPrice': 3500, 'modalPrice': 3500}, {'arrivalDate': 1490918400000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1491091200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300} ] weekly_dataframe = pd.DataFrame(datewise_prices) weekly_dataframe.drop('minPrice', axis=1, inplace=True) weekly_dataframe.drop('maxPrice', axis=1, inplace=True) weekly_dataframe['arrivalDate'] = pd.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pd.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') ``` Transferred from @hussaintamboli 's input. It will help if you provide more information, for example, the full trace when it error. @BranYang and @TomAugspurger, I have added the stacktrace. Just figured it out myself. It's because of numpy==1.13.0. For numpy==1.12.1, it's working fine. this looks fine. what exactly is the issue? On staging environment with the dependencies mentioned in `Output of pd.show_versions() on Staging`, I get `TypeError: data type "datetime" not understood`. But on local environment with dependencies mentioned in `Output of pd.show_versions() on local`, I get desired output. your example does not repro. I would guess that you have strings (and not datetimes) in one of your fields Reproduction steps. This works ``` $ pip freeze | grep numpy numpy==1.12.1 import pandas as pd datewise_prices = [ {'arrivalDate': 1490227200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490400000000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1490832000000, 'maxPrice': 3500, 'minPrice': 3500, 'modalPrice': 3500}, {'arrivalDate': 1490918400000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300}, {'arrivalDate': 1491091200000, 'maxPrice': 3300, 'minPrice': 3300, 'modalPrice': 3300} ] weekly_dataframe = pd.DataFrame(datewise_prices) weekly_dataframe.drop('minPrice', axis=1, inplace=True) weekly_dataframe.drop('maxPrice', axis=1, inplace=True) weekly_dataframe['arrivalDate'] = pd.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pd.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') weekly_dataframe.drop('arrivalDate', axis=1, inplace=True) print weekly_dataframe.to_dict(orient='records') [{'label': u'20 Mar to 26 Mar', 'modalPrice': 3300.0}, {'label': u'27 Mar to 02 Apr', 'modalPrice': 3366.6666666666665}] ``` On the staging environment where it gives error ``` pip install numpy==1.13.0 weekly_dataframe = pd.DataFrame(datewise_prices) weekly_dataframe.drop('minPrice', axis=1, inplace=True) weekly_dataframe.drop('maxPrice', axis=1, inplace=True) weekly_dataframe['arrivalDate'] = pd.to_datetime(weekly_dataframe['arrivalDate'], unit='ms') weekly_dataframe = weekly_dataframe.resample('W', on='arrivalDate')['modalPrice'].mean().dropna().reset_index() weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pd.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') TypeError: data type "datetime" not understood ``` this works on master for both numpies so not sure what your issue is I can reproduce it on 0.19.2 / 1.13.0. Small reproducible example: ``` pd.Series(pd.date_range("2012-01-01", periods=3)) - pd.offsets.DateOffset(days=6) ``` and the error actually comes from ``` np.dtype('M8[ns]') in ('datetime', 'datetime64') ``` But on master / 0.20.2, this does not raise anymore, so I suppose we have already fixed this (this line of code is also removed). Nonetheless, always welcome to do a PR to add a test for this specific case (the small reproducible example I showed above), to make sure it keeps on working. yeah we had some refactorings of this in 0.20.0. ok test PR it is. @hussaintamboli Would you be interested to add such a test? Sure. I'll do it.

2017-06-21T11:24:41Z

[]

Traceback (most recent call last): File "/root/myapp/myapp/handlers/aggregation_handler.py", line 290, in get_daily_weekly_and_monthly_aggregates weekly_dataframe['label'] = (weekly_dataframe['arrivalDate'] - pandas.offsets.DateOffset(days=6)).dt.strftime('%d %b') + ' to ' + weekly_dataframe['arrivalDate'].dt.strftime('%d %b') File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 696, in wrapper converted = _Op.get_op(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 330, in get_op return _TimeOp(left, right, name, na_op) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 343, in __init__ lvalues = self._convert_to_array(left, name=name) File "/usr/local/lib/python2.7/dist-packages/pandas/core/ops.py", line 454, in _convert_to_array if (inferred_type in ('datetime64', 'datetime', 'date', 'time') or TypeError: data type "datetime" not understood

11,263

pandas-dev/pandas

pandas-dev__pandas-16926

a587d568d213c62307a72d98d6913239f55844e8

diff --git a/doc/source/whatsnew/v0.21.0.txt b/doc/source/whatsnew/v0.21.0.txt --- a/doc/source/whatsnew/v0.21.0.txt +++ b/doc/source/whatsnew/v0.21.0.txt @@ -162,6 +162,7 @@ I/O - Bug in :func:`read_csv` in which non integer values for the header argument generated an unhelpful / unrelated error message (:issue:`16338`) +- Bug in :func:`read_stata` where value labels could not be read when using an iterator (:issue:`16923`) Plotting ^^^^^^^^ diff --git a/pandas/io/stata.py b/pandas/io/stata.py --- a/pandas/io/stata.py +++ b/pandas/io/stata.py @@ -997,6 +997,7 @@ def __init__(self, path_or_buf, convert_dates=True, self.path_or_buf = BytesIO(contents) self._read_header() + self._setup_dtype() def __enter__(self): """ enter context manager """ @@ -1299,6 +1300,23 @@ def _read_old_header(self, first_char): # necessary data to continue parsing self.data_location = self.path_or_buf.tell() + def _setup_dtype(self): + """Map between numpy and state dtypes""" + if self._dtype is not None: + return self._dtype + + dtype = [] # Convert struct data types to numpy data type + for i, typ in enumerate(self.typlist): + if typ in self.NUMPY_TYPE_MAP: + dtype.append(('s' + str(i), self.byteorder + + self.NUMPY_TYPE_MAP[typ])) + else: + dtype.append(('s' + str(i), 'S' + str(typ))) + dtype = np.dtype(dtype) + self._dtype = dtype + + return self._dtype + def _calcsize(self, fmt): return (type(fmt) is int and fmt or struct.calcsize(self.byteorder + fmt)) @@ -1472,22 +1490,10 @@ def read(self, nrows=None, convert_dates=None, if nrows is None: nrows = self.nobs - if (self.format_version >= 117) and (self._dtype is None): + if (self.format_version >= 117) and (not self._value_labels_read): self._can_read_value_labels = True self._read_strls() - # Setup the dtype. - if self._dtype is None: - dtype = [] # Convert struct data types to numpy data type - for i, typ in enumerate(self.typlist): - if typ in self.NUMPY_TYPE_MAP: - dtype.append(('s' + str(i), self.byteorder + - self.NUMPY_TYPE_MAP[typ])) - else: - dtype.append(('s' + str(i), 'S' + str(typ))) - dtype = np.dtype(dtype) - self._dtype = dtype - # Read data dtype = self._dtype max_read_len = (self.nobs - self._lines_read) * dtype.itemsize @@ -1958,7 +1964,6 @@ def _prepare_categoricals(self, data): return data get_base_missing_value = StataMissingValue.get_base_missing_value - index = data.index data_formatted = [] for col, col_is_cat in zip(data, is_cat): if col_is_cat: @@ -1981,8 +1986,7 @@ def _prepare_categoricals(self, data): # Replace missing values with Stata missing value for type values[values == -1] = get_base_missing_value(dtype) - data_formatted.append((col, values, index)) - + data_formatted.append((col, values)) else: data_formatted.append((col, data[col])) return DataFrame.from_items(data_formatted)

Unable to read Stata value_labels from .dta-file created by pandas (to_stata()) #### Code Sample, a copy-pastable example if possible ```python import pandas as pd d = {'A':['B','E','C','A','E']} df = pd.DataFrame(data=d) df['A'] = df['A'].astype('category') # Setting as categorical, similar to value_label in Stata df.to_stata('test.dta') # Writing dataframe to Stata-file dfs_fromstata = pd.read_stata('test.dta', iterator=True) # Creating StataReader-object print(dfs_fromstata.value_labels()) # Printing value_labels ``` ```python Traceback (most recent call last): File "<stdin>", line 1, in <module> print(dfs_fromstata.value_labels()) # Printing value_labels File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1725, in value_labels self._read_value_labels() File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1329, in _read_value_labels offset = self.nobs * self._dtype.itemsize AttributeError: 'NoneType' object has no attribute 'itemsize' ``` #### Problem description It seems as if read_stata() is not able to read value_labels properly from a .dta-file created by to_stata(). If the file is created by Stata itself, the value_labels are read correctly. Also, if the .dta-file created by to_stata() is opened and saved by Stata, the value_labels are read correctly. #### Expected Output {'A': {0: 'A', 1: 'B', 2: 'C', 3: 'E'}} #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 3.6.0.final.0 python-bits: 64 OS: Windows OS-release: 10 machine: AMD64 processor: Intel64 Family 6 Model 78 Stepping 3, GenuineIntel byteorder: little LC_ALL: None LANG: None LOCALE: None.None pandas: 0.20.2 pytest: 3.0.5 pip: 9.0.1 setuptools: 27.2.0 Cython: 0.25.2 numpy: 1.12.1 scipy: 0.19.1 xarray: None IPython: 5.1.0 sphinx: 1.5.1 patsy: 0.4.1 dateutil: 2.6.0 pytz: 2016.10 blosc: None bottleneck: 1.2.1 tables: 3.2.2 numexpr: 2.6.2 feather: None matplotlib: 2.0.2 openpyxl: 2.4.1 xlrd: 1.0.0 xlwt: 1.2.0 xlsxwriter: 0.9.6 lxml: 3.7.2 bs4: 4.5.3 html5lib: None sqlalchemy: 1.1.5 pymysql: None psycopg2: None jinja2: 2.9.4 s3fs: None pandas_gbq: None pandas_datareader: None </details>

cc @bashtage I don't think this is supported. It works fine. You need to read once in order to initialize some values. This appears to be an omission in the iterator implementation. Should probably raise with a usable error in the scenario. ``` import pandas as pd d = {'A':['B','E','C','A','E']} df = pd.DataFrame(data=d) df['A'] = df['A'].astype('category') # Setting as categorical, similar to value_label in Stata df.to_stata('test.dta') # Writing dataframe to Stata-file dfs_fromstata = pd.read_stata('test.dta', iterator=True) # Creating StataReader-object dfs_fromstata.read() print(dfs_fromstata.value_labels()) ``` pull-requests welcome! Should pull this block https://github.com/pandas-dev/pandas/blob/master/pandas/io/stata.py#L1480 out to a stand alone function and then call it once the header has been read.

2017-07-14T17:22:36Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> print(dfs_fromstata.value_labels()) # Printing value_labels File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1725, in value_labels self._read_value_labels() File "C:\Anaconda3\lib\site-packages\pandas\io\stata.py", line 1329, in _read_value_labels offset = self.nobs * self._dtype.itemsize AttributeError: 'NoneType' object has no attribute 'itemsize'

11,285

pandas-dev/pandas

pandas-dev__pandas-16930

4efe6560e07f28de6a1834fa90e31cef31b0fb18

diff --git a/doc/source/whatsnew/v0.21.0.txt b/doc/source/whatsnew/v0.21.0.txt --- a/doc/source/whatsnew/v0.21.0.txt +++ b/doc/source/whatsnew/v0.21.0.txt @@ -164,6 +164,8 @@ I/O - Bug in :func:`read_stata` where value labels could not be read when using an iterator (:issue:`16923`) +- Bug in :func:`read_html` where import check fails when run in multiple threads (:issue:`16928`) + Plotting ^^^^^^^^ - Bug in plotting methods using ``secondary_y`` and ``fontsize`` not setting secondary axis font size (:issue:`12565`) diff --git a/pandas/io/html.py b/pandas/io/html.py --- a/pandas/io/html.py +++ b/pandas/io/html.py @@ -37,8 +37,6 @@ def _importers(): if _IMPORTS: return - _IMPORTS = True - global _HAS_BS4, _HAS_LXML, _HAS_HTML5LIB try: @@ -59,6 +57,8 @@ def _importers(): except ImportError: pass + _IMPORTS = True + ############# # READ HTML #

read_html() Thread Safety #### Code Sample ```python #!/usr/bin/python3 import pandas import threading def fetch_file(): url = "https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.html" pandas.read_html(url) thread1 = threading.Thread(target = fetch_file) thread2 = threading.Thread(target = fetch_file) thread1.start() thread2.start() ``` ### Output ``` Exception in thread Thread-2: Traceback (most recent call last): File "/usr/lib/python3.6/threading.py", line 916, in _bootstrap_inner self.run() File "/usr/lib/python3.6/threading.py", line 864, in run self._target(*self._args, **self._kwargs) File "./pandas_bug.py", line 7, in fetch_file pandas.read_html(url) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 904, in read_html keep_default_na=keep_default_na) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 731, in _parse parser = _parser_dispatch(flav) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 691, in _parser_dispatch raise ImportError("lxml not found, please install it") ImportError: lxml not found, please install it ``` #### Problem description read_html() doesn't appear to be multi-threading safe. This specific issue seems to be caused by setting `_IMPORTS` in html.py to True too early resulting in the second thread entering `_parser_dispatch` and throwing an exception while the first thread hasn't finished the check. I have written a potential fix and will open a PR shortly. #### Expected Output No exception should be thrown since lxml is installed and the program works fine without multi-threading. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------------ commit: None python: 3.6.1.final.0 python-bits: 64 OS: Linux OS-release: 4.11.3-1-ARCH machine: x86_64 processor: byteorder: little LC_ALL: None LANG: en_GB.UTF-8 LOCALE: en_GB.UTF-8 pandas: 0.20.1 pytest: None pip: 9.0.1 setuptools: 36.0.1 Cython: None numpy: 1.12.1 scipy: 0.19.0 xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: 2.0.2 openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: 4.6.0 html5lib: 0.999999999 sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None pandas_gbq: None pandas_datareader: None </details>

2017-07-14T19:25:11Z

[]

Traceback (most recent call last): File "/usr/lib/python3.6/threading.py", line 916, in _bootstrap_inner self.run() File "/usr/lib/python3.6/threading.py", line 864, in run self._target(*self._args, **self._kwargs) File "./pandas_bug.py", line 7, in fetch_file pandas.read_html(url) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 904, in read_html keep_default_na=keep_default_na) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 731, in _parse parser = _parser_dispatch(flav) File "/usr/lib/python3.6/site-packages/pandas/io/html.py", line 691, in _parser_dispatch raise ImportError("lxml not found, please install it") ImportError: lxml not found, please install it

11,286

pandas-dev/pandas

pandas-dev__pandas-17169

9b07ef4a5b656a1532512c270533053ee338e30d

diff --git a/pandas/core/internals.py b/pandas/core/internals.py --- a/pandas/core/internals.py +++ b/pandas/core/internals.py @@ -323,10 +323,6 @@ def reindex_axis(self, indexer, method=None, axis=1, fill_value=None, fill_value=fill_value, mask_info=mask_info) return self.make_block(new_values, fastpath=True) - def get(self, item): - loc = self.items.get_loc(item) - return self.values[loc] - def iget(self, i): return self.values[i] @@ -1658,13 +1654,6 @@ def set(self, locs, values, check=False): assert locs.tolist() == [0] self.values = values - def get(self, item): - if self.ndim == 1: - loc = self.items.get_loc(item) - return self.values[loc] - else: - return self.values - def putmask(self, mask, new, align=True, inplace=False, axis=0, transpose=False, mgr=None): """ @@ -4722,8 +4711,6 @@ def _concat_indexes(indexes): def _block2d_to_blocknd(values, placement, shape, labels, ref_items): """ pivot to the labels shape """ - from pandas.core.internals import make_block - panel_shape = (len(placement),) + shape # TODO: lexsort depth needs to be 2!!

AttributeError Block.items https://github.com/pandas-dev/pandas/blob/master/pandas/core/internals.py#L326 `core.internals.Block` references `self.items`. AFAICT `items` is an attribute of `BlockManager`, does not exist in `Block`. ``` ser = pd.Series(range(5)) mgr = ser._data block = mgr.blocks[0] >>> block.get(1) Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python2.7/site-packages/pandas/core/internals.py", line 329, in get loc = self.items.get_loc(item) AttributeError: 'IntBlock' object has no attribute 'items' ```

I think this is dead code. We *always* use positional indexing on a block (e.g. ``.iget``). not sure why the linter doesn't find this. you can do a PR to remove. Code coverage also indicates this is never used (at least in our tests)

2017-08-04T07:13:48Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/local/lib/python2.7/site-packages/pandas/core/internals.py", line 329, in get loc = self.items.get_loc(item) AttributeError: 'IntBlock' object has no attribute 'items'

11,323

pandas-dev/pandas

pandas-dev__pandas-17194

3c833db29b6f5977c78d1ade791a09a5b29cedb8

diff --git a/doc/source/whatsnew/v0.21.0.txt b/doc/source/whatsnew/v0.21.0.txt --- a/doc/source/whatsnew/v0.21.0.txt +++ b/doc/source/whatsnew/v0.21.0.txt @@ -278,6 +278,7 @@ Indexing - Fixes bug where indexing with ``np.inf`` caused an ``OverflowError`` to be raised (:issue:`16957`) - Bug in reindexing on an empty ``CategoricalIndex`` (:issue:`16770`) - Fixes ``DataFrame.loc`` for setting with alignment and tz-aware ``DatetimeIndex`` (:issue:`16889`) +- Avoids ``IndexError`` when passing an Index or Series to ``.iloc`` with older numpy (:issue:`17193`) I/O ^^^ diff --git a/pandas/core/internals.py b/pandas/core/internals.py --- a/pandas/core/internals.py +++ b/pandas/core/internals.py @@ -861,6 +861,9 @@ def _is_empty_indexer(indexer): # set else: + if _np_version_under1p9: + # Work around GH 6168 to support old numpy + indexer = getattr(indexer, 'values', indexer) values[indexer] = value # coerce and try to infer the dtypes of the result

passing Index or Series to iloc fails with old numpy #### Code Sample, a copy-pastable example if possible ```python >>> s = pd.Series([1,2]) >>> s.iloc[pd.Series([0])] = 2 Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/indexing.py", line 198, in __setitem__ self._setitem_with_indexer(indexer, value) File "pandas/core/indexing.py", line 619, in _setitem_with_indexer value=value) File "pandas/core/internals.py", line 3313, in setitem return self.apply('setitem', **kwargs) File "pandas/core/internals.py", line 3201, in apply applied = getattr(b, f)(**kwargs) File "pandas/core/internals.py", line 864, in setitem values[indexer] = value IndexError: unsupported iterator index >>> np.version.version '1.7.0' ``` #### Problem description This is a consequence of #6168 , which is fixed in more recent numpy versions (but 1.7.0 is still a supported version). Not 100% sure this is worth fixing (i.e. how interested we are in keeping compatibility with numpy 1.8.0), but the fix is trivial, so I'll just push a PR and let you judge. #### Expected Output Just no error. #### Output of ``pd.show_versions()`` <details> INSTALLED VERSIONS ------------------ commit: None python: 2.7.3.final.0 python-bits: 64 OS: Linux OS-release: 4.9.0-3-amd64 machine: x86_64 processor: byteorder: little LC_ALL: None LANG: it_IT.UTF-8 LOCALE: None.None pandas: 0+unknown pytest: 3.2.0 pip: None setuptools: 36.2.7 Cython: None numpy: 1.7.0 scipy: None pyarrow: None xarray: None IPython: None sphinx: None patsy: None dateutil: 1.5 pytz: 2012c blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: None jinja2: None s3fs: None fastparquet: None pandas_gbq: None pandas_datareader: None >>> </details>

2017-08-07T22:07:19Z

[]

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "pandas/core/indexing.py", line 198, in __setitem__ self._setitem_with_indexer(indexer, value) File "pandas/core/indexing.py", line 619, in _setitem_with_indexer value=value) File "pandas/core/internals.py", line 3313, in setitem return self.apply('setitem', **kwargs) File "pandas/core/internals.py", line 3201, in apply applied = getattr(b, f)(**kwargs) File "pandas/core/internals.py", line 864, in setitem values[indexer] = value IndexError: unsupported iterator index

11,325

pandas-dev/pandas

pandas-dev__pandas-17201

674fb96b33c07c680844f674fcdf0767b6e3c2f9

diff --git a/doc/source/whatsnew/v0.21.1.txt b/doc/source/whatsnew/v0.21.1.txt --- a/doc/source/whatsnew/v0.21.1.txt +++ b/doc/source/whatsnew/v0.21.1.txt @@ -88,7 +88,7 @@ I/O - :func:`read_parquet` now allows to specify kwargs which are passed to the respective engine (:issue:`18216`) - Bug in parsing integer datetime-like columns with specified format in ``read_sql`` (:issue:`17855`). - Bug in :meth:`DataFrame.to_msgpack` when serializing data of the numpy.bool_ datatype (:issue:`18390`) - +- Bug in :func:`read_json` not decoding when reading line deliminted JSON from S3 (:issue:`17200`) Plotting ^^^^^^^^ diff --git a/pandas/io/json/json.py b/pandas/io/json/json.py --- a/pandas/io/json/json.py +++ b/pandas/io/json/json.py @@ -5,7 +5,7 @@ import pandas._libs.json as json from pandas._libs.tslib import iNaT -from pandas.compat import StringIO, long, u +from pandas.compat import StringIO, long, u, to_str from pandas import compat, isna from pandas import Series, DataFrame, to_datetime, MultiIndex from pandas.io.common import (get_filepath_or_buffer, _get_handle, @@ -458,8 +458,10 @@ def read(self): if self.lines and self.chunksize: obj = concat(self) elif self.lines: + + data = to_str(self.data) obj = self._get_object_parser( - self._combine_lines(self.data.split('\n')) + self._combine_lines(data.split('\n')) ) else: obj = self._get_object_parser(self.data) @@ -612,7 +614,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: dtype = np.dtype(dtype) return data.astype(dtype), True - except: + except (TypeError, ValueError): return data, False if convert_dates: @@ -628,7 +630,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: data = data.astype('float64') result = True - except: + except (TypeError, ValueError): pass if data.dtype.kind == 'f': @@ -639,7 +641,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: data = data.astype('float64') result = True - except: + except (TypeError, ValueError): pass # do't coerce 0-len data @@ -651,7 +653,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, if (new_data == data).all(): data = new_data result = True - except: + except (TypeError, ValueError): pass # coerce ints to 64 @@ -661,7 +663,7 @@ def _try_convert_data(self, name, data, use_dtypes=True, try: data = data.astype('int64') result = True - except: + except (TypeError, ValueError): pass return data, result @@ -680,7 +682,7 @@ def _try_convert_to_date(self, data): if new_data.dtype == 'object': try: new_data = data.astype('int64') - except: + except (TypeError, ValueError): pass # ignore numbers that are out of range @@ -697,7 +699,7 @@ def _try_convert_to_date(self, data): unit=date_unit) except ValueError: continue - except: + except Exception: break return new_data, True return data, False

read_json(lines=True) broken for s3 urls in Python 3 (v0.20.3) #### Code Sample, a copy-pastable example if possible Using Python ```python import pandas as pd inputdf = pd.read_json(path_or_buf="s3://path/to/python-lines/file.json", lines=True) ``` The file is similar to: ``` {"url": "blah", "other": "blah"} {"url": "blah", "other": "blah"} {"url": "blah", "other": "blah"} ``` #### Problem description When attempting to read a python lines file into a DataFrame using the s3 protocol, the above code will error with: ``` 2017-08-08 11:06:14,225 - image_rank_csv - ERROR - initial_value must be str or None, not bytes Traceback (most recent call last): File "image_rank_csv.py", line 62, in run inputdf = pd.read_json(path_or_buf="s3://path/to/python-lines/file.json", lines=True) File "...env/lib/python3.6/site-packages/pandas/io/json/json.py", line 347, in read_json lines = list(StringIO(json.strip())) TypeError: initial_value must be str or None, not bytes ``` This works fine if the file is local, e.g.: ```python import pandas as pd inputdf = pd.read_json(path_or_buf="/local/path/to/python-lines/file.json", lines=True) ``` #### Expected Output Expect to successfully read the file and error above not to occur. My current thinking is that when we get the file handle: https://github.com/pandas-dev/pandas/blob/v0.20.3/pandas/io/json/json.py#L333 , you delegate to `s3fs`, which documents that [it only operates in Binary mode](http://s3fs.readthedocs.io/en/latest/#limitations). Therefore when you `read()`: https://github.com/pandas-dev/pandas/blob/v0.20.3/pandas/io/json/json.py#L335, Therefore passing to `StringIO` will fail here: https://github.com/pandas-dev/pandas/blob/v0.20.3/pandas/io/json/json.py#L347 . Maybe it needs a different handler for `BytesIO`? #### Output of ``pd.show_versions()`` <details> [paste the output of ``pd.show_versions()`` here below this line] ``` INSTALLED VERSIONS ------------------ commit: None python: 3.6.1.final.0 python-bits: 64 OS: Darwin OS-release: 16.6.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 pandas: 0.20.3 pytest: None pip: 9.0.1 setuptools: 36.2.7 Cython: None numpy: 1.12.0 scipy: 0.19.1 xarray: None IPython: None sphinx: None patsy: None dateutil: 2.6.0 pytz: 2017.2 blosc: None bottleneck: None tables: None numexpr: None feather: None matplotlib: None openpyxl: None xlrd: None xlwt: None xlsxwriter: None lxml: None bs4: None html5lib: None sqlalchemy: None pymysql: None psycopg2: 2.6.2 (dt dec pq3 ext lo64) jinja2: None s3fs: 0.1.2 pandas_gbq: None pandas_datareader: None ``` </details>

Well we do have a `BytesIO` class in `pandas.compat`. If we can condition on the data returned to us, that would be reasonable I think. @gfyoung I'm not intimately familiar with the codebase, but I have a (possibly naive) fix that just attempts to decode the json based on whatever `encoding` is. It seems to fix. Would you like to review if do a PR? > I'm not intimately familiar with the codebase Don't worry, that's a pretty tall order :smile: Absolutely! Submit a PR, and we'll certainly review it. FYI, here's how we handle it on the CSV side: https://github.com/pandas-dev/pandas/blob/3c833db29b6f5977c78d1ade791a09a5b29cedb8/pandas/io/common.py#L401 that `f` would be an instance of `S3File`, which is one of the classes in `need_text_wrapping`. By the `TextIOWrapper` can go around buffer of bytes and it'll do the encoding. You might be able to reuse parts of that for `read_json`, or just do something similar.

2017-08-08T18:39:45Z

[]

Traceback (most recent call last): File "image_rank_csv.py", line 62, in run inputdf = pd.read_json(path_or_buf="s3://path/to/python-lines/file.json", lines=True) File "...env/lib/python3.6/site-packages/pandas/io/json/json.py", line 347, in read_json lines = list(StringIO(json.strip())) TypeError: initial_value must be str or None, not bytes

11,326