Datasets:

andfanilo

/

streamlit-issues

like 1

https://api.github.com/repos/streamlit/streamlit/issues/6136

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6136/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6136/comments

https://api.github.com/repos/streamlit/streamlit/issues/6136/events

https://github.com/streamlit/streamlit/issues/6136

I_kwDODCoeTs5eicmX

Protobuff version upgrade

open

NONE

### Problem Streamlit requires Protobuf version < 4. Many modern libraries need > 4. This results in collision when managing project dependencies. ### Solution Move to newer Protobuf. Would be great to know if this can be done!

https://api.github.com/repos/streamlit/streamlit/issues/6136/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6135

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6135/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6135/comments

https://api.github.com/repos/streamlit/streamlit/issues/6135/events

https://github.com/streamlit/streamlit/pull/6135

PR_kwDODCoeTs5KCYa4

Improve data editor and dataframe docstring

open

COLLABORATOR

## 📚 Context Improve data editor and dataframe docstring for `width` and `height`. --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6135/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6134

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6134/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6134/comments

https://api.github.com/repos/streamlit/streamlit/issues/6134/events

https://github.com/streamlit/streamlit/issues/6134

I_kwDODCoeTs5ehlts

Histogram slider

open

NONE

### Problem A user interacting with a slider might be interested in the distribution of the values in order to more appropriately filter relevant data. As far as I am aware, this is not yet possible. ### Solution **MVP:** Use a histogram to show the distribution of each step/bin. **Possible additions:** Move the range bar sideways, coloring steps/bins inside of the range. ### Additional context For inspiration, it could look something like the [vue-histogram-slider](https://github.com/oguzhaninan/vue-histogram-slider): <p align="center"> <img src="https://github.com/oguzhaninan/vue-histogram-slider/raw/master/resources/histogram-slider-demo.gif"> </p>

https://api.github.com/repos/streamlit/streamlit/issues/6134/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6133

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6133/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6133/comments

https://api.github.com/repos/streamlit/streamlit/issues/6133/events

https://github.com/streamlit/streamlit/issues/6133

I_kwDODCoeTs5ehZFI

Image gallery

open

NONE

Any plans to add image gallery? Could be quite useful (and popular) since its the era of diffusion models

https://api.github.com/repos/streamlit/streamlit/issues/6133/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6132

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6132/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6132/comments

https://api.github.com/repos/streamlit/streamlit/issues/6132/events

https://github.com/streamlit/streamlit/issues/6132

I_kwDODCoeTs5egY0S

Horizontal layout for fixed-with elements

open

NONE

### Problem Quite often one would like to have multiple elements such as `st.button` next to each other. There is currently no ideal solution to do this. Columns are an acceptable workaround, but elements such as `st.button` or `st.metric` do not fill the entire column (correctly), leaving large gaps between elements. Tricks like using `st.columns([1,1,10])` to lay out two buttons and a bunch of whitespace can be unpredictable. ### Solution **MVP:** A container-type command, like `st.horizontal_array` or some such, which would allow for sequential horizontal layout of fixed-width elements such as `st.button`, `st.metric` side-by-side. **Possible additions:** A consideration of tweaking the gap (like with `st.column`) and what happens in cases of overflow (multi-line sequence?). ### Additional context Not an unpopular request on the forums: https://discuss.streamlit.io/t/st-button-in-one-line/25966 https://discuss.streamlit.io/t/two-buttons-on-the-same-line/2749 https://discuss.streamlit.io/t/multiple-download-buttons-on-the-same-line/37706

https://api.github.com/repos/streamlit/streamlit/issues/6132/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6131

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6131/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6131/comments

https://api.github.com/repos/streamlit/streamlit/issues/6131/events

https://github.com/streamlit/streamlit/issues/6131

I_kwDODCoeTs5ee6d1

allow_output_mutation does not exist anymore

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary Verified that "st.cache" is deprecated. No similar examples of document references were found. ### Reproducible Code Example ```Python @st.cache(allow_output_mutation=True) def cached_model(): model = SentenceTransformer('jhgan/ko-sroberta-multitask') return model @st.cache(allow_output_mutation=True) def get_dataset(): df = pd.read_csv('wellness_dataset.csv') df['embedding'] = df['embedding'].apply(json.loads) return df ``` ### Steps To Reproduce 1. @st.cache_resource(allow_output_mutation=True) def cached_model(): model = SentenceTransformer('jhgan/ko-sroberta-multitask') return model @st.cache_resource(allow_output_mutation=True) def get_dataset(): df = pd.read_csv('wellness_dataset.csv') df['embedding'] = df['embedding'].apply(json.loads) return df ### Expected Behavior Page not found. ### Current Behavior Please check the official document of st.cache. ### Is this a regression? - [x] Yes, this used to work in a previous version. ### Debug info <font class="papago-parent"><font class="papago-source" style="display:none;">- Streamlit version:</font>- 스트림라이트 버전:</font><font class="papago-parent"><font class="papago-source" style="display:none;"> </font> </font><font class="papago-parent"><font class="papago-source" style="display:none;">- Python version:</font>- Python 버전:</font><font class="papago-parent"><font class="papago-source" style="display:none;"> </font> </font><font class="papago-parent"><font class="papago-source" style="display:none;">- Operating System:</font>- 운영체제:</font><font class="papago-parent"><font class="papago-source" style="display:none;"> </font> </font><font class="papago-parent"><font class="papago-source" style="display:none;">- Browser:</font>- 브라우저:</font><font class="papago-parent"><font class="papago-source" style="display:none;"> </font> </font><font class="papago-parent"><font class="papago-source" style="display:none;">- Virtual environment:</font>- 가상 환경:</font><font class="papago-parent"><font class="papago-source" style="display:none;"> </font> </font> ### Additional Information _No response_ ### Are you willing to submit a PR? - [X] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6131/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6130

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6130/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6130/comments

https://api.github.com/repos/streamlit/streamlit/issues/6130/events

https://github.com/streamlit/streamlit/issues/6130

I_kwDODCoeTs5eeVbr

Copy Column Headers from Streamlit Dataframe

open

NONE

### Problem Streamlit not copying column headers when selecting and copying data is a borderline no-go for using the tool as our datasets are quite wide in some cases and without headers the overhead is very painful to get useful data into excel/JMP/jupyter. We don't use this often as streamlit is mostly a dashboard-y tool, easily cutting data is a huge bonus. Having the default include headers or optionally being able to configure headers on copy would be very positive. ### Solution Minium Suggestion: ```python st.dataframe(df) # default copy behavior is copying data and columns ``` **Possible additions:** Ideally we could choose (though 90% of the time we want both) ``` python #default config works like it does now not copying column headers but copying data def dataframe(df,copy_headers=False,copy_data=True): ``` This allows us to write code like this: ``` python st.dataframe(df) # Backwards compatible default state st.dataframe(df,copy_headers=False,copy_data=True) # Explicity config to default state. st.dataframe(df,copy_headers=True,copy_data=True) # Header st.dataframe(df,copy_headers=True,copy_data=False) # Columns only ???... ``` Finally another useful...but remotely so, would be ``` python st.dataframe(df,copy_schema=True) # return data types of each column perhaps int/str/datetime/float ``` I'm currently look at providing buttons with each data frame that use data frame code to save or download data based on solutions I've seen in here and on SO.

https://api.github.com/repos/streamlit/streamlit/issues/6130/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6129

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6129/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6129/comments

https://api.github.com/repos/streamlit/streamlit/issues/6129/events

https://github.com/streamlit/streamlit/pull/6129

PR_kwDODCoeTs5J-Zqm

Fix `st.metric` tooltip position regression

open

CONTRIBUTOR

## 📚 Context This PR fixes a regression in reported by @snehankekre, where the `help` tooltip in `st.metric` would show up in the wrong place. - What kind of change does this PR introduce? - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - Added `display: flex;` to `StyleTruncateText` so we can place the `(?)` icon next to the label; - Added `overflow: hidden;` to the `StyledStreamlitMarkdown` label container; - Added `overflow: hidden;` and `text-overflow: ellipsis;` to the actual `<p>` tag inside the label container. - [ ] This is a breaking API change - [x] This is a visible (user-facing) change **Revised:**   **Current:**  ## 🧪 Testing Done - [x] Screenshots included - [ ] Added/Updated unit tests - [x] Added/Updated e2e tests ## 🌐 References - **Issue**: Closes #6093 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6129/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6128

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6128/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6128/comments

https://api.github.com/repos/streamlit/streamlit/issues/6128/events

https://github.com/streamlit/streamlit/pull/6128

PR_kwDODCoeTs5J-WtT

cache_errors: don't refer to experimental_memo|singleton

open

COLLABORATOR

Our `UnserializableReturnValueError` cache error currently refers to our old cache decorator names (`@st.experimental_memo|singleton`), and our old experimental cache decorator docs. This PR updates the error text to refer to the new decorators and the new cache docs.

https://api.github.com/repos/streamlit/streamlit/issues/6128/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6127

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6127/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6127/comments

https://api.github.com/repos/streamlit/streamlit/issues/6127/events

https://github.com/streamlit/streamlit/pull/6127

PR_kwDODCoeTs5J-Vnj

Fix detecting arrow df as widget when not editable

closed

COLLABORATOR

The data editor introduced a new kind of state for an element, being a widget only conditionally. Element replay code relied on the presence of an id field in the protobuf for determining what is a widget, but arrow dfs don't act like a widget, and don't call `register_widget`, unless used with the data editor. In this situation, the id isn't actually set on the protobuf, so we can use it being set for our widget detection. Fixes #6103 ## 📚 Context - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - [ ] This is a breaking API change - [x] This is a visible (user-facing) change ## 🧪 Testing Done - [ ] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References - **Issue**: #6103

https://api.github.com/repos/streamlit/streamlit/issues/6127/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6126

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6126/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6126/comments

https://api.github.com/repos/streamlit/streamlit/issues/6126/events

https://github.com/streamlit/streamlit/issues/6126

I_kwDODCoeTs5eczl0

Lottie animations and eCharts gauge disappear when moving from one tab to another

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary I have a program with five tabs. In Chrome, Firefox, and Edge as you move from tab to tab, the lottie animation and eCharts gauge work in the tab you are in, but they both disappear in the previous tabs. However this does not happen in Safari. ### Reproducible Code Example ```Python import streamlit as st import requests, json, time from PIL import Image from streamlit_lottie import st_lottie from streamlit_echarts import st_echarts st.set_page_config(page_title="Test Lottie and Gauge", page_icon=Image.open('images/favicon.ico'), layout="wide") govt = ['file://images/govt1.json','file://images/govt2.json'] ## def load_lottie_resource(resource: str): if resource[:4] == 'http': r = requests.get(resource) if r.status_code !=200: return None else: return r.json() elif resource[:4] == 'file': with open(resource[7:], 'r') as f: return json.load(f) ## def gov_total(df): st.info('Testing') st_lottie(govt_animation[1],key=time.time(),height=120,width=120,speed=0.7) option = { "tooltip": {"formatter": '{a} <br/>{b} : {c}%'}, "series": [{ "name": 'Test', "type": 'gauge', "radius": '90%', "data": [{ "value": df, "name": 'test' }], "progress": { "show": "true" }, "pointer": { "length": '50%', "valueAnimation": "true", "width": 5, "offsetCenter": [0, '3%'] }, "detail": { "valueAnimation": "true", "formatter": '{value}%', "offsetCenter": [0, '70%'], }, }] } st_echarts(options=option, renderer="svg", key=time.time()) return ##MAIN govt_animation = list(map(load_lottie_resource,govt)) ## -- navigation setup tab1, tab2, tab3, tab4, tab5 = st.tabs(['Tab1',':hospital: Tab2',':hospital: Tab3',':globe_with_meridians: Tab4',':round_pushpin: Tab5']) with tab1: st.write('tab1') gov_total(51) with tab2: st.write('tab2') gov_total(52) with tab3: st.write('tab3') gov_total(53) with tab4: st.write('tab4') gov_total(54) with tab5: st.write('tab5') gov_total(55) ``` ### Steps To Reproduce Click thru the tabs and return to a previous tab and the lottie animation and eCharts gauge will disappear ### Expected Behavior the lottie animation and eChart gauge should remain as it does in the Safari broswer ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version:1.170, 1.18.0, 1.18.1 - Python version:10.8 - Operating System:Ubuntu 22, Mac OS 13.1 - Browser: Chrome, Firefox, Edge (all latest) - Virtual environment: Conda ### Additional Information Any pointers would be greatly appreciated ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6126/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6125

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6125/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6125/comments

https://api.github.com/repos/streamlit/streamlit/issues/6125/events

https://github.com/streamlit/streamlit/issues/6125

I_kwDODCoeTs5ecV-C

st_lottie animations and echarts gauges disappear on previous tabs when you move to another tab

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary I have a Steamlit app that when I change tabs and use a selectbox to make a selection and the page(s) re-render, the st_lottie animation and echart gauge disappear on previous tabs. I have tested this on Streamlit version 1.17.0 and 1.18.0 and I get the same results. Lottie version 0.0.3, echarts version 0.4.0. This happens consistently in Chrome but works correctly in Safari. What is even more strange is that it works perfectly on Safari on Streamlit 1.17.0, however if I upgrade to Streamlit 1.18.0 I get a blank page in Safari. ### Reproducible Code Example ```Python def draw_lottie_and_gauge(df): st.info('Government program coverage') st_lottie(govt_animation[1],key=time.time(),height=120,width=300,speed=0.7) option = { "tooltip": {"formatter": '{a} <br/>{b} : {c}%'}, "series": [{ "name": 'Government Programs', "type": 'gauge', "radius": '90%', "data": [{ "value": round(df.loc[(df['Payor'] == 'Total Government')]['Percent'].sum() * 100), "name": 'government' }], "progress": { "show": "true" }, "pointer": { "length": '50%', "valueAnimation": "true", "width": 5, "offsetCenter": [0, '3%'] }, "detail": { "valueAnimation": "true", "formatter": '{value}%', "offsetCenter": [0, '70%'], }, }] } st_echarts(options=option, renderer="svg", key=time.time()) return ``` ### Steps To Reproduce Click on a tab select a criteria from the selectbox. dashboard information generates, including the Lottie animation and echart gauge go to a previous tab and everything is populated **except** the lottie animation and echarts gauge ### Expected Behavior The lottie animation and echarts gauge should remain on each tab ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version:1.17.0 and 1.18.0 - Python version:3.10.8 - Operating System:Ubuntu 22 and Mac OSx 13.1 - Browser: Chrome - Virtual environment: Conda If Streamlit Version 1.17.0 everything works in Safari and does not work in Chrome. If Streamlit Version 1.18.0 Safari shows a blank page. Page shows up in Chrome but the lottie animation and echarts gauge disappear just like in 1.17.0 ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6125/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6124

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6124/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6124/comments

https://api.github.com/repos/streamlit/streamlit/issues/6124/events

https://github.com/streamlit/streamlit/issues/6124

I_kwDODCoeTs5ebqin

Favicon on Streamlit page does not show up

closed

NONE

### Summary When configuring a page icon using `set_page_config`, the icon does not show up on tab of the Streamlit page. ### Reproducible Code Example ```Python import streamlit as st st.set_page_config(page_title="Test page", page_icon="📝", layout="wide") st.write("The page icon should be 📝") ``` ### Steps To Reproduce 1. Run the Reproducible Code Example 2. Observe the tab icon in a browser ### Expected Behavior The tab should show the icon given by `page_icon="📝"`. ### Current Behavior The tab shows an empty page as icon. ### Is this a regression? I don't know, since it never worked since I started using Streamlit. ### Debug info - Streamlit version: 1.11.0 - Python version: 3.8.10 - Operating System: Windows 10 Enterprise, Version 21H2, OS build 19044.2486 - Browser: Microsoft Edge Version 110.0.1587.41 (64-Bit) ### Additional Information While diving into this issue, I found that the icon given above resolve to the following line in the HTML of the Streamlit page: ``` <link rel="shortcut icon" href="https://twemoji.maxcdn.com/2/72x72/1f4dd.png"> ``` When entering this address directly into the browser, the file cannot be found. This seems to be due to an update within the folder structure of twemoji. A working link would be ``` https://twemoji.maxcdn.com/2/svg/1f4dd.svg ``` ### Are you willing to submit a PR? - [X] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6124/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6123

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6123/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6123/comments

https://api.github.com/repos/streamlit/streamlit/issues/6123/events

https://github.com/streamlit/streamlit/pull/6123

PR_kwDODCoeTs5J8s4u

Pin unixodbc version to fix CI pipeline

closed

COLLABORATOR

## 📚 Context The CI pipeline is broken because it tries to install a newer version of `unixodbc-dev` (2.3.11). It fails to update the old version that already exists in the container (`2.3.9-5`).     - What kind of change does this PR introduce? - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - As a quick fix, we pin the `unixodbc-dev` dependency to the old version. --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6123/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6122

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6122/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6122/comments

https://api.github.com/repos/streamlit/streamlit/issues/6122/events

https://github.com/streamlit/streamlit/pull/6122

PR_kwDODCoeTs5J8h-I

Add editing support for numeric index types

closed

COLLABORATOR

## 📚 Context The current implementation does not allow the deprecated numeric index types in Pandas: `Int64Index`, `Float64Index` and `UInt64Index`. Even so, they are deprecated, there isn't any specific reason why we should not support them when used with older Pandas versions. - What kind of change does this PR introduce? - [ ] Bugfix - [x] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - Allow numeric indices as well. We need to check based on the string type name, since they are planned to be removed from pandas in future versions. This would potentially break if we do an instance check. ## 🧪 Testing Done - [ ] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6122/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6121

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6121/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6121/comments

https://api.github.com/repos/streamlit/streamlit/issues/6121/events

https://github.com/streamlit/streamlit/issues/6121

I_kwDODCoeTs5ebFRx

A specific polygon value is not showing color and is just showing the value.

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary One particular Polygon is not showing any color but when you hover over it, it's showing the value with the particular ICS Code it refers to. All the other polygon values are showing color and are being plotted without any problem. I want to fix this issue so that all polygons show color in the map. ### Reproducible Code Example ```Python import streamlit as st import json import geopandas as gpd import pyproj import plotly.graph_objs as go # FLU Full Shape File Read # reading in the points shapefile FLU_Full_map_df = gpd.read_file(data_dir_shape + f'FLU_Full.shp') # project GeoPandas dataframe # set GeoJSON file path FLU_Full_map_df.to_file(data_dir_shape + f'geojson_FLU_Full.json', driver = "GeoJSON") with open(data_dir_shape + f'geojson_FLU_Full.json') as geofile1: j_file_FLU_Full = json.load(geofile1) # index geojson i = 0 for feature in j_file_FLU_Full["features"]: feature['id'] = str(i).zfill(1) i += 1 # FLU Short Shape File Read # reading in the points shapefile FLU_Short_map_df = gpd.read_file(data_dir_shape + f'FLU_Short.shp') # project GeoPandas dataframe # set GeoJSON file path FLU_Short_map_df.to_file(data_dir_shape + f'geojson_FLU_Short.json', driver = "GeoJSON") with open(data_dir_shape + f'geojson_FLU_Short.json') as geofile1: j_file_FLU_Short = json.load(geofile1) # index geojson i = 0 for feature in j_file_FLU_Short["features"]: feature['id'] = str(i).zfill(1) i += 1 # Typhoid Full Shape File Read # reading in the points shapefile Typhoid_Full_map_df = gpd.read_file(data_dir_shape + f'Typhoid_Full.shp') # project GeoPandas dataframe # set GeoJSON file path Typhoid_Full_map_df.to_file(data_dir_shape + f'geojson_Typhoid_Full.json', driver = "GeoJSON") with open(data_dir_shape + f'geojson_Typhoid_Full.json') as geofile1: j_file_Typhoid_Full = json.load(geofile1) # index geojson i = 0 for feature in j_file_Typhoid_Full["features"]: feature['id'] = str(i).zfill(1) i += 1 # Hepatitis A Full Shape File Read # reading in the points shapefile Hepatitis_A_Full_map_df = gpd.read_file(data_dir_shape + f'Hepatitis_A_Full.shp') # project GeoPandas dataframe # set GeoJSON file path Hepatitis_A_Full_map_df.to_file(data_dir_shape + f'geojson_Hepatitis_A_Full.json', driver = "GeoJSON") with open(data_dir_shape + f'geojson_Hepatitis_A_Full.json') as geofile1: j_file_Hepatitis_A_Full = json.load(geofile1) # index geojson i = 0 for feature in j_file_Hepatitis_A_Full["features"]: feature['id'] = str(i).zfill(1) i += 1 # Hepatitis A Full Shape File Read # reading in the points shapefile Yellow_fever_Full_map_df = gpd.read_file(data_dir_shape + f'Yellow_fever_Full.shp') # project GeoPandas dataframe # set GeoJSON file path Yellow_fever_Full_map_df.to_file(data_dir_shape + f'geojson_Yellow_fever_Full.json', driver = "GeoJSON") with open(data_dir_shape + f'geojson_Yellow_fever_Full.json') as geofile1: j_file_Yellow_fever_Full = json.load(geofile1) # index geojson i = 0 for feature in j_file_Yellow_fever_Full["features"]: feature['id'] = str(i).zfill(2) i += 1 print(FLU_Short_map_df['NIC_Per_Pe']) # mapbox token mapboxtoken = 'MAPBOXTOKEN' # define layers and plot map FLU_Full = go.Choroplethmapbox(z = FLU_Full_map_df['NIC_Per_Pe'], locations = [x['id'] for x in j_file_FLU_Full['features']], colorscale='Cividis', geojson=j_file_FLU_Full, text = FLU_Full_map_df['ICS_CODE']) # define layers and plot map FLU_Short = go.Choroplethmapbox(z = FLU_Short_map_df['NIC_Per_Pe'], locations = [x['id'] for x in j_file_FLU_Short['features']], colorscale='Cividis', geojson=j_file_FLU_Short, text = FLU_Short_map_df['ICS_CODE']) # define layers and plot map Typhoid_Full = go.Choroplethmapbox(z = Typhoid_Full_map_df['NIC_Per_Pe'], locations = [x['id'] for x in j_file_Typhoid_Full['features']], colorscale='Cividis', geojson=j_file_Typhoid_Full, text = Typhoid_Full_map_df['ICS_CODE']) # define layers and plot map Hepatitis_A_Full = go.Choroplethmapbox(z = Hepatitis_A_Full_map_df['NIC_Per_Pe'], locations = [x['id'] for x in j_file_Hepatitis_A_Full['features']], colorscale='Cividis', geojson=j_file_Hepatitis_A_Full, text = Hepatitis_A_Full_map_df['ICS_CODE']) # define layers and plot map Yellow_fever_Full = go.Choroplethmapbox(z = Yellow_fever_Full_map_df['NIC_Per_Pe'], locations = [x['id'] for x in j_file_Yellow_fever_Full['features']], colorscale='Cividis', geojson=j_file_Yellow_fever_Full, text = Yellow_fever_Full_map_df['ICS_CODE']) layout = go.Layout(title_text = 'Map of Total NIC Per Person', width = 800, height = 700, mapbox = dict(center = dict(lat=54.5, lon=-3), accesstoken=mapboxtoken, style="light", zoom=4.5)) activity = ['FLU_Full', 'FLU_Short', 'Typhoid_Full', 'Hepatitis_A_Full', 'Yellow_fever_Full'] choice = st.sidebar.selectbox('Drug Selection: ', activity) if choice == 'FLU_Full': fig = go.Figure(data=FLU_Full, layout=layout) # display streamlit map elif choice == 'FLU_Short': fig = go.Figure(data=FLU_Short, layout=layout) elif choice == 'Typhoid_Full': fig = go.Figure(data=Typhoid_Full, layout=layout) elif choice == 'Hepatitis_A_Full': fig = go.Figure(data=Hepatitis_A_Full, layout=layout) else: fig = go.Figure(data=Yellow_fever_Full, layout=layout) st.title('NIC Per Person: ') st.plotly_chart(fig) ``` ### Steps To Reproduce If you run this code after generating a mapboxtoken and run it with streamlit, it's showing something like this: <img width="1399" alt="Screenshot 2023-02-14 at 13 17 14" src="https://user-images.githubusercontent.com/84929423/218750000-7e574c31-99f2-4f7e-94d9-bc0889f5f9f9.png"> ### Expected Behavior <img width="1399" alt="Screenshot 2023-02-14 at 13 17 47" src="https://user-images.githubusercontent.com/84929423/218750110-2ed8bbc3-804a-4e1f-96c5-bd97cc859647.png"> If you see, I want that white spot where Lincolnshire is to be filled with color. All the other counties in UK are being filled with color. ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: - Python version: - Operating System: - Browser: - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6121/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6120

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6120/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6120/comments

https://api.github.com/repos/streamlit/streamlit/issues/6120/events

https://github.com/streamlit/streamlit/issues/6120

I_kwDODCoeTs5eZzzb

Unable to reboot streamlit app deployed on streamlit server. It says I have only view only access

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary I am trying to reboot my streamlit app https://mehulgupta2016154-resume-builder-streamlit-app-ajmqjx.streamlit.app/ but when trying the reboot button, it says I have view-only access and cant reboot. Please help ### Reproducible Code Example _No response_ ### Steps To Reproduce _No response_ ### Expected Behavior _No response_ ### Current Behavior _No response_ ### Is this a regression? - [X] Yes, this used to work in a previous version. ### Debug info - Streamlit version: - Python version: - Operating System: - Browser: - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6120/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6119

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6119/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6119/comments

https://api.github.com/repos/streamlit/streamlit/issues/6119/events

https://github.com/streamlit/streamlit/issues/6119

I_kwDODCoeTs5eZt2R

Slider impacts button state

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary I have an `if` statement that depends on the state of a button. When there is no slider inside the `if` statement, everything works fine. ``` import streamlit as st if st.button("Validate"): st.write("Hello") else: st.write("Bye") ``` When I add a slider within the if-statement, once I use the slider, the code in the else part is executed. ``` import streamlit as st if st.button("Validate"): st.write("Hello") st.slider("Slider example", min_value=0, max_value=10) else: st.write("Bye") ``` I don't know why the slider impacts the state of the button (as the `else` statement is executed) ### Reproducible Code Example ```Python import streamlit as st if st.button("Validate"): st.write("Hello") st.slider("Slider example", min_value=0, max_value=10) else: st.write("Bye") ``` ### Steps To Reproduce _No response_ ### Expected Behavior The code within the `else` statement shouldn't be executed when one uses the slider that is within the `if`. ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.18.1 - Python version: 3.7.13 - Operating System: Mac OS Ventura 13.2 ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6119/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6118

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6118/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6118/comments

https://api.github.com/repos/streamlit/streamlit/issues/6118/events

https://github.com/streamlit/streamlit/issues/6118

I_kwDODCoeTs5eZJS-

bug sit

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary  ### Reproducible Code Example _No response_ ### Steps To Reproduce _No response_ ### Expected Behavior _No response_ ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: - Python version: - Operating System: - Browser: - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6118/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6117

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6117/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6117/comments

https://api.github.com/repos/streamlit/streamlit/issues/6117/events

https://github.com/streamlit/streamlit/issues/6117

I_kwDODCoeTs5eYigN

Streamlit Cloud should support the --no-root option for poetry install

open

CONTRIBUTOR

### Problem When using Poetry with a project that does not contain a main package, Streamlit Cloud fails to install the dependencies. An example project structure that fails is below. ``` some-project ├── LICENSE ├── README.md ├── app.py ├── poetry.lock └── pyproject.toml ``` It is because `poetry install` emits the following error. ```shell $ poetry install (...) Installing dependencies from lock file (...) /path/to/some-project/some_project does not contain any element ``` The command `poetry install` expects this project contains a package root like `some-project/some_project/__init__.py` and tries to install the package itself in addition to its dependencies, while it actually doesn't exist. ### Solution `poetry install` provide an option to deal with this case, `--no-root`. (See https://python-poetry.org/docs/cli/#install) I think Streamlit Cloud should run `poetry install` with the `--no-root` option because usually we don't need to deal with the Streamlit app projects as packages to install, but we just want to install the dependencies managed with Poetry. ### Additional context [This repo](https://github.com/whitphx/streamlit-webrtc-hugging-face-transformer-example) is an actual project with which I encountered this problem. [This revision](https://github.com/whitphx/streamlit-webrtc-hugging-face-transformer-example/tree/a6b77c2aacd182c7bb8409b725021aca1c6f7b6e) of the repo caused this error, and I fixed it by adding `streamlit_webrtc_hugging_face_transformer_example/__init__.py` as a workaround in [this commit](https://github.com/whitphx/streamlit-webrtc-hugging-face-transformer-example/commit/237026ca4b19dcb6e9b94240cfe1adb0657946fe) so that it works on Streamlit Cloud now. --- The same error was reported at [this thread in the forum](https://discuss.streamlit.io/t/streamlit-cloud-deploy-is-failing-with-error-does-not-contain-any-element/31541) --- Community voting on feature requests enables the Streamlit team to understand which features are most important to our users. **If you'd like the Streamlit team to prioritize this feature request, please use the 👍 (thumbs up emoji) reaction in response to the initial post.**

https://api.github.com/repos/streamlit/streamlit/issues/6117/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6116

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6116/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6116/comments

https://api.github.com/repos/streamlit/streamlit/issues/6116/events

https://github.com/streamlit/streamlit/issues/6116

I_kwDODCoeTs5eYPH9

[Icon cannot display] Couldn't find the requested file xxx.png in twemoji.

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary When running a steamlit server, the icon cannot fetch (404).  ### Reproducible Code Example ```Python import streamlit as st if __name__ == "__main__": st.set_page_config( page_title="wocao", page_icon="random", layout="wide", initial_sidebar_state="expanded", ) x = st.slider("Select a value") st.write(x, "squared is", x * x) ``` ### Steps To Reproduce streamlit run demo.py ### Expected Behavior Show the icon I selected or random ### Current Behavior Cannot show the icon but a blank icon  ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.17.0 - Python version: 3.7.12 - Operating System: Ubuntu 20 - Browser: Edge - Virtual environment: Conda ### Additional Information _No response_ ### Are you willing to submit a PR? - [X] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6116/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6115

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6115/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6115/comments

https://api.github.com/repos/streamlit/streamlit/issues/6115/events

https://github.com/streamlit/streamlit/issues/6115

I_kwDODCoeTs5eX_TK

Issue with Deploy app

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary Dear Streamlit I have 'stream_app.py' on GitHub.com https://github.com/JohnTan38/Telco-Customer-Churn-Prediction/blob/main/stream_app.py This file 'model_C_1.bin' is on GutHub.com too https://github.com/JohnTan38/Telco-Customer-Churn-Prediction/blob/main/model_C_1.bin Code snippet: model_file = (r"https://raw.githubusercontent.com/JohnTan38/Telco-Customer-Churn-Prediction/blob/main/model_C_1.bin") with open(model_file, 'rb') as f_in: dv, model = pickle.load(f_in) Here's the error message when I tried to deploy s'tream_app.py' -------------------------------------------------------------------------------------------------------------------------------------------- File "/home/appuser/venv/lib/python3.9/site-packages/streamlit/runtime/scriptrunner/script_runner.py", line 565, in _run_script exec(code, module.__dict__) File "/app/telco-customer-churn-prediction/stream_app.py", line 7, in <module> with open(model_file, 'rb') as f_in: -------------------------------------------------------------------------------------------------------------------------------------------- ### Reproducible Code Example ```Python model_file = (r"https://raw.githubusercontent.com/JohnTan38/Telco-Customer-Churn-Prediction/blob/main/model_C_1.bin") with open(model_file, 'rb') as f_in: dv, model = pickle.load(f_in) ``` ### Steps To Reproduce _No response_ ### Expected Behavior A page opens for user input Contact email psg6868@yahoo.com azx1818@hotmail.com ### Current Behavior _No response_ ### Is this a regression? - [X] Yes, this used to work in a previous version. ### Debug info - Streamlit version: - Python version: - Operating System: - Browser: - Virtual environment: ---------------------- Python 3.9 Windows 10 Chrome Browser ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6115/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6114

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6114/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6114/comments

https://api.github.com/repos/streamlit/streamlit/issues/6114/events

https://github.com/streamlit/streamlit/pull/6114

PR_kwDODCoeTs5J5tsV

Fix member function caching

closed

COLLABORATOR

We have a regression in caching - starting in 1.18, using `@st.cache_data` (or `@st.cache_resource`) on a class member function will fail with an error when the cached function was called: ```python class TestClass: @cache_decorator def foo(_self): # We underscore-prefix `_self`, because our class is not hashable. return "it works!" obj = TestClass() result = obj.foo() # Explode! ``` This is the result of a refactor that moved "cached function wrapper" logic into a class, `CachedFunc`, (from a closure). `functools.update_wrapper` (and `functools.wraps`) has unexpected behavior when applied to a class member function: the wrapped function no longer passes the "self" argument through to the underlying member function. I don't know why this is! I couldn't find anything in the `functools` docs (or in a quick Google search) to explain the behavior. My guess is that using `update_wrapper` on a class instance ends up overwriting one or more instance properties that break its member function call semantics. But the fix is simple: we hoist the `functools.wraps` call outside of `CachedFunc`, rather than using `functools` to modify the `CachedFunc` member function itself. Fixes https://github.com/streamlit/streamlit/issues/6109

https://api.github.com/repos/streamlit/streamlit/issues/6114/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6113

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6113/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6113/comments

https://api.github.com/repos/streamlit/streamlit/issues/6113/events

https://github.com/streamlit/streamlit/pull/6113

PR_kwDODCoeTs5J5Onq

Improve logging when arrow serialization fails

closed

COLLABORATOR

## 📚 Context We currently print out the full stack trace as an info log when the Arrow serialization fails. This was added #6022 as debugging help during development but isn't that useful for the users. This PR removes the stack trace and only shows the Arrow exception message. - What kind of change does this PR introduce? - [ ] Bugfix - [ ] Feature - [x] Refactoring - [ ] Other, please describe: --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6113/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6112

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6112/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6112/comments

https://api.github.com/repos/streamlit/streamlit/issues/6112/events

https://github.com/streamlit/streamlit/issues/6112

I_kwDODCoeTs5eWkwR

st.date_input monday as start of week

closed

NONE

widget st.date_input don't have option to view Monday as first day of week.

https://api.github.com/repos/streamlit/streamlit/issues/6112/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6111

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6111/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6111/comments

https://api.github.com/repos/streamlit/streamlit/issues/6111/events

https://github.com/streamlit/streamlit/pull/6111

PR_kwDODCoeTs5J36se

Use initial widget value in case data editor was removed

closed

COLLABORATOR

## 📚 Context In the current version, as soon as the `data_editor` widget is removed it will lose all state (e.g. edits). This also happens when another element is added above the `data_editor`. Other widgets handle this situation by initializing some of the state from the widget state. We need to do the same with the data editor. - What kind of change does this PR introduce? - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - Split up the `useDataLoader` into two hooks: One hook for loading the columns and one for handling the cell data. - Save widget state whenever the value changes (and not just on trigger reload) - Initialize from the widget state JSON if the widget state already exist in the widget state manager. ## 🧪 Testing Done - [ ] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6111/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6110

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6110/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6110/comments

https://api.github.com/repos/streamlit/streamlit/issues/6110/events

https://github.com/streamlit/streamlit/pull/6110

PR_kwDODCoeTs5J3vo7

[WIP] Revert Metric help changes for now

closed

COLLABORATOR

<!-- Before contributing (PLEASE READ!) ⚠️ If your contribution is more than a few lines of code, then prior to starting to code on it please post in the issue saying you want to volunteer, then wait for a positive response. And if there is no issue for it yet, create it first. This helps make sure: 1. Two people aren't working on the same thing 2. This is something Streamlit's maintainers believe should be implemented/fixed 3. Any API, UI, or deeper architectural changes that need to be implemented have been fully thought through by Streamlit's maintainers 4. Your time is well spent! More information in our wiki: https://github.com/streamlit/streamlit/wiki/Contributing --> ## 📚 Context _Please describe the project or issue background here_ - What kind of change does this PR introduce? - [ ] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - _Add bullet points summarizing your changes here_ - [ ] This is a breaking API change - [ ] This is a visible (user-facing) change **Revised:** _Insert screenshot of your updated UI/code here_ **Current:** _Insert screenshot of existing UI/code here_ ## 🧪 Testing Done - [ ] Screenshots included - [ ] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: Closes #XXXX --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6110/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6109

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6109/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6109/comments

https://api.github.com/repos/streamlit/streamlit/issues/6109/events

https://github.com/streamlit/streamlit/issues/6109

I_kwDODCoeTs5eTJ3z

No longer able to st.cache_data instance methods

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary Previously using experimental_memo the following worked fine. After deprecation and changing to cache_data, the methos isn't called with underscored parameters anymore. Interestingly, using experimental_memo ALSO fails now. ### Reproducible Code Example ```Python import streamlit as st class MyView: def __init__(self, txt): self.txt = txt self.view() def view(self): st.write(self.txt) st.write(self.load_data()) @st.cache_data(ttl=60) def load_data(_self): return _self.txt MyView("OH YES") ``` ### Steps To Reproduce streamlit run code.py ### Expected Behavior Working ### Current Behavior TypeError: load_data() missing 1 required positional argument: '_self' ### Is this a regression? - [X] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.18.1 (working with 1.14.0) - Python version: - Operating System: - Browser: - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6109/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6108

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6108/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6108/comments

https://api.github.com/repos/streamlit/streamlit/issues/6108/events

https://github.com/streamlit/streamlit/issues/6108

I_kwDODCoeTs5eS5YG

Pre-cache on start-up

open

NONE

The `st.cache` is really useful (I'm still on 1.13) but I think its user-friendliness can be improved. Now, the first run still takes a long time for the first user. But it would be useful if we can "pre-cache" the result of a function so that even the first user does not have to wait too long for long running processes, e.g. for downloading a ML model behind the scenes. A solution would be to add an option to st.cache like "precache" that accepts a list of dicts, where each dict is a kwarg for the function. On app start up, the function is then run for all those different types of kwargs. It's just an idea but I think it could be useful!

https://api.github.com/repos/streamlit/streamlit/issues/6108/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6107

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6107/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6107/comments

https://api.github.com/repos/streamlit/streamlit/issues/6107/events

https://github.com/streamlit/streamlit/issues/6107

I_kwDODCoeTs5eRuTV

date_input on sidebar has a display issue

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary The `st.sidebar.date_input()` has a display issue when the sidebar is too thin. It cuts the left part of the calendar widget. ### Reproducible Code Example ```Python import streamlit as st st.sidebar.date_input('Foo') ``` ### Steps To Reproduce 1. Create the date_input on the sidebar; 2. With the cursor, set the sidebar to a thinner width; 3. Click on the date_input and see the calendar's left side cut off ### Expected Behavior I expect the calendar to adjust its position when the sidebar width is thinner than the calendar width. ### Current Behavior It cuts the left part of the calendar widget. ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: Streamlit, version 1.17.0 - Python version: Python 3.10.6 - Operating System: Linux Pop Os - Browser: Firefox & Chrome - Virtual environment: Standard Python Env ### Additional Information I've checked that on both my monitors (19 & 24 inch) and both Firefox and Chrome browsers. Here are the images showing the display issue:   ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6107/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6106

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6106/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6106/comments

https://api.github.com/repos/streamlit/streamlit/issues/6106/events

https://github.com/streamlit/streamlit/issues/6106

I_kwDODCoeTs5eQPVN

Add preview under each PDF file in st.file_uploader with accept_multiple_files=True

closed

NONE

### Problem When using file_uploader with accept_multiple_files=True, a list of uploaded files is being displayed. However, when PDF preview is added using displayPDF, the previews are displayed below this list while missing the relation of each preview to its file. ### Solution Add the preview under each file in the file list

https://api.github.com/repos/streamlit/streamlit/issues/6106/timeline

not_planned

https://api.github.com/repos/streamlit/streamlit/issues/6105

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6105/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6105/comments

https://api.github.com/repos/streamlit/streamlit/issues/6105/events

https://github.com/streamlit/streamlit/issues/6105

I_kwDODCoeTs5eP64d

Update images in table to st.cache_data instead of st.experimental_memo

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary When I use images in table extra, I get this error message 'st.experimental_memo is deprecated. Please use the new command st.cache_data instead, which has the same behavior. More information [in our docs](https://docs.streamlit.io/library/advanced-features/caching).' just need to update the source code with st.cache_data ### Reproducible Code Example _No response_ ### Steps To Reproduce _No response_ ### Expected Behavior _No response_ ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: - Python version: - Operating System: - Browser: - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6105/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6104

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6104/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6104/comments

https://api.github.com/repos/streamlit/streamlit/issues/6104/events

https://github.com/streamlit/streamlit/issues/6104

I_kwDODCoeTs5eOjg8

Audio widget not working on iOS Chrome/Safari and Safari

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary It seems that st.audio() stopped working on iOS and Safari. Similar [issue](https://github.com/streamlit/streamlit/issues/2113) was already resolved in 2020 so…potential regression?  issue was confirmed by other streamlit users: https://discuss.streamlit.io/t/st-audio-stopped-working-on-ios-safari-again/37049 ### Reproducible Code Example _No response_ ### Steps To Reproduce 1. Go to https://adventure.streamlit.app/ 2. Enter your name 3. Check if audio is working (it is on windows and android :) ) 4. Github repo: https://github.com/tomjohnh/streamlit-game/blob/main/streamlit_app.py ### Expected Behavior Audio widget is working ### Current Behavior Audio widget is not working ("Error" message) ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.17 - Python version: - Operating System: iOS - Browser: Chrome on iOS / Safari - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6104/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6103

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6103/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6103/comments

https://api.github.com/repos/streamlit/streamlit/issues/6103/events

https://github.com/streamlit/streamlit/issues/6103

I_kwDODCoeTs5eOcHe

st.cache_resource (v 1.18) vs st.experimental_singleton (v 1.17): attributeerror with cache_resource

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary AttributeError Traceback: File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py", line 565, in _run_script exec(code, module.__dict__) File "C:\Users\A260219\Jupyter\bloomberg\prova.py", line 2126, in <module> ottimizzazione(num_variabili, fun_varianza, ptf_weights_equally, expected_returns, lista_asset_pairplot) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cache_utils.py", line 186, in __call__ return self._get_or_create_cached_value(args, kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cache_utils.py", line 211, in _get_or_create_cached_value return self._handle_cache_miss(cache, value_key, func_args, func_kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cache_utils.py", line 265, in _handle_cache_miss computed_value = self._info.func(*func_args, **func_kwargs) File "C:\Users\A260219\Jupyter\bloomberg\prova.py", line 1555, in ottimizzazione st.write(pd.DataFrame(data=np.round(ptf_weights.reshape([1, num_variabili])*100, decimals=1), columns=dati_ritorni.columns, index=["Pesi"])) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\metrics_util.py", line 311, in wrapped_func result = non_optional_func(*args, **kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\elements\write.py", line 200, in write self.dg.dataframe(arg) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\metrics_util.py", line 311, in wrapped_func result = non_optional_func(*args, **kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\elements\dataframe_selector.py", line 116, in dataframe return self.dg._arrow_dataframe( File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\metrics_util.py", line 311, in wrapped_func result = non_optional_func(*args, **kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\elements\arrow.py", line 117, in _arrow_dataframe return self.dg._enqueue("arrow_data_frame", proto) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\delta_generator.py", line 563, in _enqueue caching.save_element_message( File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\__init__.py", line 50, in save_element_message CACHE_RESOURCE_MESSAGE_REPLAY_CTX.save_element_message( File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cached_message_replay.py", line 293, in save_element_message raise AttributeError ### Reproducible Code Example ```Python @st.cache_resource ##al momento non funzionante con la funzione di cui di seguito e la versione 1.18 di Streamlit (ma st.experimental_singleton funzionava nella 1.17) def ottimizzazione(num_variabili: int, _fun_varianza: Callable[[np.array], float], ptf_weights_equally: np.array, expected_returns: np.array, lista_asset_pairplot: list) -> None: #l'underscore serviva per dire a streamlit di non fare l'hash della funzione fun_varianza=_fun_varianza #qui si specificano i vincoli (min, max) per le singole variabili (NB: il numero di vincoli deve essere esattamente uguale al numero di variabili) bnds = tuple([(valore_minimo_per_singola_asset_class, valore_massimo_per_singola_asset_class) for i in range(num_variabili)]) st.write("--------------------------------------------------------------------") st.write("Elaborazione portafoglio di minima varianza") #di seguito la chiamata all'ottimizzatore con argomenti: la funzione-oggetto da minimizzare, X0 ovvero un initial guess delle variabili, il metodo scelto ('SLSQP' é #l'unico metodo che supporta ogni tipo di vincolo), i bounds sui valori delle singole variabili, i vincoli su piú variabili simultaneamente, infine le opzioni opt={'maxiter': maxiter, 'disp': True, 'ftol': 1e-12, 'eps': 1.4901161193847656e-08} #valori di default 'eps': 1.4901161193847656e-08, 'ftol': 1e-06 (si tratta rispettivamente della step size (eps) usata per l'approssimazione numerica della jacobiana ovvero il vettore del gradiente, e dell'obbiettivo di precisione (ftol) sulla funzione obbiettivo) res = minimize(fun_varianza, x0=ptf_weights_equally, method='SLSQP', bounds=bnds, constraints=st.session_state.vincoli_generali, options=opt) ptf_weights=res.x #attribuisce al vettore pesi i pesi ottimali calcolati da minimize (relativi al portafoglio di minima varianza) st.write("Di seguito i pesi degli asset in portafoglio espressi in percentuale:") st.write(pd.DataFrame(data=np.round(ptf_weights.reshape([1, num_variabili])*100, decimals=1), columns=dati_ritorni.columns, index=["Pesi"])) ptf_min_var_return=(ptf_weights.T).dot(expected_returns) #calcola il rendimento del portafoglio ottimale di minima varianza ptf_min_var_variance=res.fun #attribuisce ad apposita variabile il valore finale minimizzato della funzione obbiettivo (ovvero la varianza dei rendimenti di portafoglio) st.write("La deviazione standard annualizzata percentuale del portafoglio é %0.03f" % (np.sqrt(ptf_min_var_variance)*np.sqrt(periodicity)*100)) st.write("Il rendimento percentuale del portafoglio é %0.03f" % (ptf_min_var_return*100)) ##di seguito si calcola il portafoglio di massimo rendimento (da fare per il caso in cui vi siano vincoli complessi) #qui si specifica la funzione da minimizzare per il portafoglio di massimo rendimento def fun_rendimento(ptf_weights: np.array) -> float: return -(ptf_weights.T).dot(expected_returns) #si minimizza l'opposto del rendimento per massimizzarlo st.write("--------------------------------------------------------------------") st.write("Elaborazione portafoglio di massimo rendimento") #di seguito la massimizzazione del rendimento (minimizza l'opposto) con vincoli invariati res_return = minimize(fun_rendimento, x0=ptf_weights_equally, method='SLSQP', bounds=bnds, constraints=st.session_state.vincoli_generali, options=opt) #estrae il valore finale del rendimento (massimizato); ovviamente va invertito di segno perché la funzione obbiettivo é l'opposto del rendimento ptf_max_risk_return=-res_return.fun st.write("Di seguito i pesi degli asset in portafoglio espressi in percentuale:") st.write(pd.DataFrame(data=np.round((res_return.x).reshape([1, num_variabili])*100, decimals=1), columns=dati_ritorni.columns, index=["Pesi"])) st.write("Il rendimento percentuale del portafoglio é %0.03f" % (ptf_max_risk_return*100)) step=(ptf_max_risk_return-ptf_min_var_return)/n_step #si specifica la matrice dei pesi di portafoglio con shape variabili*portafogli ptf_weights_matrix=np.zeros([num_variabili, (n_step+1)]) ptf_weights_matrix[:, 0]=ptf_weights #il primo portafoglio é quello di minima varianza di cui giá abbiamo calcolato i pesi #si specifica il vettore delle varianze dei vari portafogli sulla frontiera efficiente variance_vector=np.zeros([(n_step+1),]) variance_vector[0]=ptf_min_var_variance #si specifica il vettore dei rendimenti dei vari portafogli sulla frontiera efficiente returns_vector=np.zeros([(n_step+1),]) returns_vector[0]=ptf_min_var_return desired_return=ptf_min_var_return+step #la funzione destinata ad essere usata per il vincolo sul rendimento nella costruzione #del resto della frontiera def vincolo_rendimento(ptf_weights: np.array) -> float: return (ptf_weights.T).dot(expected_returns) - desired_return for i in range(n_step): st.write("--------------------------------------------------------------------") st.write("Elaborazione portafoglio n.ro %s" % (i+1)) #qui si specificano i vincoli (generali, ovvero su piú variabili) nella forma di un dictionary che contiene la specifica del tipo ('eq' per equality ed #'ineq' per inequality) e la funzione del vincolo (che verrá posta uguale a zero per i vincoli di equality e non-negativa per i vincoli di inequality) res = minimize(fun_varianza, x0=ptf_weights, method='SLSQP', bounds=bnds, constraints=[i for i in st.session_state.vincoli_generali] + [{'type': 'eq', 'fun': vincolo_rendimento}], #aggiunge il vincolo sul rendimento desiderato options=opt) ptf_weights=res.x ptf_weights_matrix[:, i+1]=ptf_weights st.write("Di seguito i pesi degli asset in portafoglio espressi in percentuale:") st.write(pd.DataFrame(data=np.round(ptf_weights.reshape([1, num_variabili])*100, decimals=1), columns=dati_ritorni.columns, index=["Pesi"])) variance_vector[i+1]=res.fun st.write("La deviazione standard annualizzata percentuale del portafoglio é %0.03f" % (np.sqrt(variance_vector[i+1])*np.sqrt(periodicity)*100)) returns_vector[i+1]=(ptf_weights.T).dot(expected_returns) st.write("Il rendimento percentuale del portafoglio é %0.03f" % (returns_vector[i+1]*100)) desired_return=desired_return+step st.write("---------------------------------------------------------------------------------------------") st.write("A seguire un riepilogo della composizione dei portafogli componenti la frontiera di Markowitz:") sintesi_composizione_portafogli_markowitz=pd.DataFrame(data=np.round(ptf_weights_matrix*100, decimals=2), #i pesi in portafoglio columns=[("Ptf n.ro "+str(i)) for i in range(n_step+1)], index=dati_ritorni.columns, dtype="float") sintesi_composizione_portafogli_markowitz.to_csv("sintesi_composizione_portafogli_markowitz.csv", sep=";", decimal=".") st.write(sintesi_composizione_portafogli_markowitz) st.write("Di seguito un riepilogo di rendimenti, rischio ed efficienza dei portafogli componenti la frontiera di Markowitz:") sintesi_rendimento_rischio_portafogli_markowitz=pd.DataFrame(data=np.round([returns_vector*100, #i rendimenti np.sqrt(variance_vector)*np.sqrt(periodicity)*100, #le deviazioni standard annualizzate np.sqrt(variance_vector)*np.sqrt(periodicity)*2.566*100, #il VaR parametrico t-Student 6df con livello di confidenza 99%, orizzonte annuale, espresso in percentuale np.divide((returns_vector*100-risk_free_rate*100), np.sqrt(variance_vector)*np.sqrt(periodicity)*100)], #gli sharpe ratio decimals=2), columns=[("Ptf n.ro "+str(i)) for i in range(n_step+1)], index=["Rendimenti (%)", "Deviazioni Standard (%)", "VaR 99% t-Student 6df (%)", "Sharpe Ratio"], dtype="float") sintesi_rendimento_rischio_portafogli_markowitz.to_csv("sintesi_rendimento_rischio_portafogli_markowitz.csv", sep=";", decimal=".") st.write(sintesi_rendimento_rischio_portafogli_markowitz) #concateniamo i dati essenziali della frontiera di Markowitz, inframezzando una riga vuota e le intestazioni di colonna prima della parte rendimento-rischio empty_row=pd.Series([None for _ in range(len(sintesi_composizione_portafogli_markowitz.columns.values.tolist()))], index=sintesi_composizione_portafogli_markowitz.columns.values.tolist(), name=" ") sintesi_composizione_portafogli_markowitz_format=sintesi_composizione_portafogli_markowitz.append(empty_row, ignore_index=False) intestazione_row=pd.Series([f"Ptf n.ro {i}" for i in range(len(sintesi_composizione_portafogli_markowitz.columns.values.tolist()))], index=sintesi_composizione_portafogli_markowitz.columns.values.tolist(), name=" ") sintesi_composizione_portafogli_markowitz_format=sintesi_composizione_portafogli_markowitz_format.append(intestazione_row, ignore_index=False) st.session_state.riassuntivo_markowitz=pd.concat([sintesi_composizione_portafogli_markowitz_format, sintesi_rendimento_rischio_portafogli_markowitz], axis=0) #plotto il grafico con la composizione della frontiera di Markowitz fig=px.area(sintesi_composizione_portafogli_markowitz.T, \ x=sintesi_composizione_portafogli_markowitz.T.index, \ y=sintesi_composizione_portafogli_markowitz.T.columns, \ labels = dict(index = "N.ro Portafoglio", value = "Percentuale in portafoglio"), \ title="Andamento composizione frontiera di Markowitz") st.plotly_chart(fig, use_container_width=True) #calcolo le deviazioni standard annualizzate std_dev_annualized_vector=np.sqrt(variance_vector)*np.sqrt(periodicity) #plotto la frontiera di Markowitz fig=px.line(x=std_dev_annualized_vector*100, \ y=returns_vector*100, \ labels = dict(x = "Deviazione standard annualizzata di portafoglio (%)", y = "Rendimento annuale (%)"), \ title="Frontiera di Markowitz") st.plotly_chart(fig, use_container_width=True) st.write("Di seguito la matrice di correlazione:") #calcolo la matrice di correlazione df_corr=pd.DataFrame.corr(pd.DataFrame(data=returns, columns=dati_ritorni.columns), method='pearson') #plotto la heatmap fig=px.imshow(df_corr) st.plotly_chart(fig, use_container_width=True) if plotting_grafico_pairplot=="y": st.write("Di seguito il grafico pairplot:") #la riga di codice seguente elimina eventuali indici che fossero superiori al consentito (ad esempio 12 o 8 su un numero di variabili/asset pari a 8) lista_asset_pairplot=[lista_asset_pairplot[i] for i in range(len(lista_asset_pairplot)) if lista_asset_pairplot[i]<num_variabili] df_ritorni=pd.DataFrame(data=returns, columns=dati_ritorni.columns) df_ritorni=df_ritorni.loc[:, [dati_ritorni.columns[i] for i in lista_asset_pairplot]] ##per debug #st.write(df_ritorni) #creo uno scatter plot fig_scatter = px.scatter_matrix(df_ritorni, width=1500, height=1000) #cambiamo dimensione ed opacitá dei punti fig_scatter.update_traces(marker={"size": 4,"opacity":0.7}) #stampo a schermo lo scatterplot (nota bene il mancato uso del metodo show) fig_scatter if uso_resampled_frontiers: ##di seguito il codice per il calcolo delle resampled frontiers #crea un vettore numpy contenente le medie attese delle variabili, ovvero in questo caso i rendimenti periodicizzati (ovvero trasformati da annuali in settimanali o mensili) means=(1+expected_returns)**(1/periodicity)-1 #esempio di come si possa vettorializzare il calcolo (velocizzandolo) #inizializza la matrice 3d che contiene i pesi di tutte le frontiere efficienti ptf_weights_3darray=np.zeros([num_variabili, (n_step+1), n_scenari]) #inizializza le matrici destinate a contenere varianze e ritorni (da resampling) delle varie frontiere efficienti variance_vector_2darray=np.zeros([(n_step+1), n_scenari]) returns_vector_2darray=np.zeros([(n_step+1), n_scenari]) #inizializza le matrici destinate a contenere di volta in volta le covarianze ed i #rendimenti ricampionati dei singoli scenari covariance_matrix_sampling=np.zeros([num_variabili, num_variabili]) returns_sampling=np.zeros([num_variabili]) #crea matrice vuota n*num_variabili destinata ad accogliere i returns periodali ricampionati con t-Student multivariata output_t=np.empty([n, num_variabili]) if plottare_andamenti_simulati=="y": #crea matrice vuota n*num_variabili*n_scenari destinata alla storicizzazione dei dati contenuti nella matrice di cui sopra output_t_3d=np.empty([n, num_variabili, n_scenari]) #qui si specifica la funzione da minimizzare che deve essere espressa come funzione di un array numpy #notare che ora si usa la covarianza calcolata sui rendimenti campionati, ovvero covariance_matrix_sampling def fun_varianza_resampling(ptf_weights: np.array) -> float: return ((ptf_weights.T).dot(covariance_matrix_sampling)).dot(ptf_weights) #w*VAR*w ovvero la varianza di portafoglio #notare che si usa returns_sampling in luogo di expected_returns (perché si usano i rendimenti dello scenario) def vincolo_rendimento_resampling(ptf_weights: np.array) -> float: return (ptf_weights.T).dot(returns_sampling) - desired_return #qui si specifica la funzione da minimizzare per il portafoglio di massimo rendimento #notare che si usa return_sampling ovvero i rendimenti ricampionati in luogo degli expected returns originali def fun_rendimento_resampling(ptf_weights: np.array) -> float: return -(ptf_weights.T).dot(returns_sampling) #si minimizza l'opposto del rendimento per massimizzarlo #inizializza le grandezze destinate a tenere traccia di eventuali anomalie nell'ottimizzatore di scipy (minimize) anomalies_count=0 anomalies_status=[] ##da qui si parte con la generazione degli scenari (via simulazioni monte carlo) e successivo calcolo delle relative ##frontiere efficienti #annuncio ed inizializzo barra avanzamento per le simulazioni st.write("\n") st.write("\n") st.write("\n") st.write("\n") st.write("\n") st.write("Avanzamento elaborazione scenari e relative frontiere ricampionate (Resampled Frontiers)") barra_avanzamento_simulazioni = st.progress(0) st.write("\n") for z in range(n_scenari): #aggiorno barra avanzamento simulazioni barra_avanzamento_simulazioni.progress(float((z+1)/n_scenari)) #genera n estrazioni (vettori) da una gaussiana multivariata con medie nulle e covarianza covariance_matrix #(ergo una matrice n*num_variabili, ovvero estrazioni*variabili) output_n=np.random.multivariate_normal(np.zeros(num_variabili), covariance_matrix, n) ##codice seguente é necessario se si vuole trasformare l'estrazione di cui sopra in una estrazione da t-student multivariata #degree of freedom di una t-student multivariata da cui si vuole estrarre campioni df=6 #n estrazioni da distribuzione chi quadro che servono per il fattore di correzione del vettore dell'estrazione dalla gaussiana w=np.random.chisquare(df, n) #attenzione che w é un vettore di dimensione n, come del resto il fattore_correzione a seguire #il fattore di correzione é pari a radice_quadrata(df/estrazione_chi_q) fattore_correzione=np.sqrt(w/df) #np.sqrt funziona element-wise fattore_correzione=fattore_correzione**(-1) #inverte per ottenere radice_quadrata(df/estrazione_chi_q) ovvero il fattore di correzione da moltiplicare per l'estrazione dalla normale #a seguire la moltiplicazione di ciascuna estrazione dalla gaussiana per il fattore di correzione corrispondente (element-wise) for i in range(n): output_t[i, :]=np.multiply(output_n[i, :], fattore_correzione[i]) #np.multiply funziona element-wise #occorre risommare il vettore delle medie in quanto le estrazioni sono fatte da distribuzioni con medie nulle output_n[i, :]=means+output_n[i, :] #nel caso della t-student si normalizza anche per la deviazione standard della t-student standardizzata (che NON é 1 come invece nel caso della gaussiana standardizzata) output_t[i, :]=means+output_t[i, :]/(np.sqrt(df/(df-2))) #serve quindi per avere una matrice covarianza finale (calcolata sul sample multivariato estratto) asintoticamente uguale alla matrice covariance_matrix originale if plottare_andamenti_simulati=="y": #storicizza i rendimenti ricampionati con t-Student ad uso successivo (grafici) output_t_3d[:, :, z]=output_t #calcola i nuovi rendimenti medi (dello scenario z) returns_sampling=np.mean(output_t, axis=0) #calcola la nuova matrice covarianza a valere sui rendimenti da scenario z covariance_matrix_sampling=np.cov(output_t, rowvar=False) #riannualizza i rendimenti returns_sampling=(1+returns_sampling)**(periodicity)-1 #qui si specificano i vincoli (min, max) per le singole variabili (NB: il numero di vincoli deve essere esattamente uguale al numero di variabili) bnds = tuple([(valore_minimo_per_singola_asset_class, valore_massimo_per_singola_asset_class) for i in range(num_variabili)]) #di seguito la chiamata all'ottimizzatore con argomenti: la funzione-oggetto da minimizzare, X0 ovvero un initial guess delle variabili, il metodo scelto ('SLSQP' é #l'unico metodo che supporta ogni tipo di vincolo), i bounds sui valori delle singole variabili, i vincoli su piú variabili simultaneamente, infine le opzioni opt={'maxiter': maxiter, 'disp': False, 'ftol': 1e-12, 'eps': 1.4901161193847656e-08} #valori di default 'eps': 1.4901161193847656e-08, 'ftol': 1e-06 (si tratta rispettivamente della step size (eps) usata per l'approssimazione numerica della jacobiana ovvero il vettore del gradiente, e dell'obbiettivo di precisione (ftol) sulla funzione obbiettivo) res = minimize(fun_varianza_resampling, x0=ptf_weights_equally, method='SLSQP', bounds=bnds, constraints=st.session_state.vincoli_generali, options=opt) #di seguito il codice che serve per tenere traccia di eventuali anomalie nell'ottimizzatore di scipy (minimize) if res.success!=True: anomalies_count+=1 anomalies_status.append(res.message) ptf_weights=res.x #attribuisce al vettore pesi i pesi ottimali calcolati da minimize (relativi al portafoglio di minima varianza) ptf_min_var_return=(ptf_weights.T).dot(returns_sampling) #calcola il rendimento del portafoglio ottimale di minima varianza ptf_min_var_variance=res.fun #attribuisce ad apposita variabile il valore finale minimizzato della funzione obbiettivo (ovvero la varianza dei rendimenti di portafoglio) ##di seguito si calcola il portafoglio di massimo rendimento (da fare per il caso in cui vi siano vincoli complessi) #di seguito la massimizzazione del rendimento (minimizza l'opposto) con vincoli invariati res_return = minimize(fun_rendimento_resampling, x0=ptf_weights_equally, method='SLSQP', bounds=bnds, constraints=st.session_state.vincoli_generali, options=opt) #di seguito il codice che serve per tenere traccia di eventuali anomalie nell'ottimizzatore di scipy (minimize) if res_return.success!=True: anomalies_count+=1 anomalies_status.append(res.message) #estrae il valore finale del rendimento (massimizato); ovviamente va invertito di segno perché la funzione obbiettivo é l'opposto del rendimento ptf_max_risk_return=-res_return.fun #come in precedenza calcola lo step di aumento dei rendimenti da aggiungere per ogni nuovo portafoglio da ottimizzare sulla frontiera step=(ptf_max_risk_return-ptf_min_var_return)/n_step ptf_weights_3darray[:, 0, z]=ptf_weights #il primo portafoglio é quello di minima varianza di cui giá abbiamo calcolato i pesi #la matrice (resampled frontiers) delle varianze con shape n_step+1, n_scenari variance_vector_2darray[0, z]=ptf_min_var_variance #la matrice (resampled frontiers) dei rendimenti con shape n_step+1, n_scenari returns_vector_2darray[0, z]=ptf_min_var_return desired_return=ptf_min_var_return+step for i in range(n_step): #si minimizza la varianza ricampionata sotto il vincolo dei rendimenti ricampionati res = minimize(fun_varianza_resampling, x0=ptf_weights, method='SLSQP', bounds=bnds, constraints=[i for i in st.session_state.vincoli_generali] + [{'type': 'eq', 'fun': vincolo_rendimento_resampling}], #aggiunge il vincolo sul rendimento desiderato options=opt) #di seguito il codice che serve per tenere traccia di eventuali anomalie nell'ottimizzatore di scipy (minimize) if res.success!=True: anomalies_count+=1 anomalies_status.append(res.message) ptf_weights=res.x ptf_weights_3darray[:, i+1, z]=ptf_weights #si salvano i pesi di portafoglio nel cubo di dimensioni num_variabili, (n_step+1), n_scenari variance_vector_2darray[i+1, z]=res.fun returns_vector_2darray[i+1, z]=(ptf_weights.T).dot(returns_sampling) desired_return=desired_return+step #si inizializza la matrice dei pesi dei portafogli sulla frontiera ricampionata ptf_weights_resampled=np.zeros([num_variabili, (n_step+1)]) #a seguire l'averaging dei pesi lungo la direzione degli scenari for j in range(n_step+1): for a in range(num_variabili): ptf_weights_resampled[a, j]=np.average(ptf_weights_3darray[a, j, :]) #si inizializzano vettori contenenti varianze e rendimenti della resampled frontier varianze_frontiera_ricampionata=np.zeros([(n_step+1)]) rendimenti_frontiera_ricampionata=np.zeros([(n_step+1)]) #qui si popolano i vettori di cui sopra #si tenga ovviamente presente che si usa varianze e rendimenti "storici" (o comunque #quelli usati nell'ottimizzazione di Markowitz) for s in range(n_step+1): varianze_frontiera_ricampionata[s]=((ptf_weights_resampled[:, s].T).dot(covariance_matrix)).dot(ptf_weights_resampled[:, s]) rendimenti_frontiera_ricampionata[s]=(ptf_weights_resampled[:, s].T).dot(expected_returns) #si passa alle deviazioni standard annualizzate dev_std_frontiera_ricampionata=(np.sqrt(varianze_frontiera_ricampionata))*np.sqrt(periodicity) st.write("---------------------------------------------------------------------------------------------") st.write("A seguire un riepilogo della composizione dei portafogli componenti la resampled frontier:") sintesi_composizione_portafogli_resampled_frontiers=pd.DataFrame(data=np.round(ptf_weights_resampled*100, decimals=2), #i pesi in portafoglio columns=[("Ptf n.ro "+str(i)) for i in range(n_step+1)], index=dati_ritorni.columns, dtype="float") sintesi_composizione_portafogli_resampled_frontiers.to_csv("sintesi_composizione_portafogli_resampled_frontiers.csv", sep=";", decimal=".") st.write(sintesi_composizione_portafogli_resampled_frontiers) st.write("Di seguito un riepilogo di rendimenti, rischio ed efficienza dei portafogli componenti la resampled frontier:") sintesi_rendimento_rischio_portafogli_resampled_frontiers=pd.DataFrame(data=np.round([rendimenti_frontiera_ricampionata*100, #i rendimenti dev_std_frontiera_ricampionata*100, #le deviazioni standard annualizzate dev_std_frontiera_ricampionata*2.566*100, #il VaR parametrico t-Student 6df con livello di confidenza 99%, orizzonte annuale, espresso in percentuale np.divide((rendimenti_frontiera_ricampionata*100-risk_free_rate*100), dev_std_frontiera_ricampionata*100)], #gli sharpe ratio decimals=2), columns=[("Ptf n.ro "+str(i)) for i in range(n_step+1)], index=["Rendimenti (%)", "Deviazioni Standard (%)", "VaR 99% t-Student 6df (%)", "Sharpe Ratio"], dtype="float") sintesi_rendimento_rischio_portafogli_resampled_frontiers.to_csv("sintesi_rendimento_rischio_portafogli_resampled_frontiers.csv", sep=";", decimal=".") st.write(sintesi_rendimento_rischio_portafogli_resampled_frontiers) #concateniamo i dati essenziali della Frontiera Ricampionata, inframezzando una riga vuota e le intestazioni di colonna prima della parte rendimento-rischio empty_row=pd.Series([None for _ in range(len(sintesi_composizione_portafogli_resampled_frontiers.columns.values.tolist()))], index=sintesi_composizione_portafogli_resampled_frontiers.columns.values.tolist(), name=" ") sintesi_composizione_portafogli_resampled_frontiers_format=sintesi_composizione_portafogli_resampled_frontiers.append(empty_row, ignore_index=False) intestazione_row=pd.Series([f"Ptf n.ro {i}" for i in range(len(sintesi_composizione_portafogli_resampled_frontiers.columns.values.tolist()))], index=sintesi_composizione_portafogli_resampled_frontiers.columns.values.tolist(), name=" ") sintesi_composizione_portafogli_resampled_frontiers_format=sintesi_composizione_portafogli_resampled_frontiers_format.append(intestazione_row, ignore_index=False) st.session_state.riassuntivo_resampled=pd.concat([sintesi_composizione_portafogli_resampled_frontiers_format, sintesi_rendimento_rischio_portafogli_resampled_frontiers], axis=0) #plotto il grafico con la composizione della Frontiera Ricampionata fig=px.area(sintesi_composizione_portafogli_resampled_frontiers.T, \ x=sintesi_composizione_portafogli_resampled_frontiers.T.index, \ y=sintesi_composizione_portafogli_resampled_frontiers.T.columns, \ labels = dict(index = "N.ro Portafoglio", value = "Percentuale in portafoglio"), \ title="Andamento composizione Frontiera Ricampionata") st.plotly_chart(fig, use_container_width=True) #calcola le deviazioni standard annualizzate std_dev_annualized_vector_resampled=np.sqrt(variance_vector_2darray)*np.sqrt(periodicity) returns_vector_resampled=returns_vector_2darray ##facciamo uno stack dei dati per il grafico a seguire ##shape degli array numpy coinvolti di seguito #std_dev_annualized_vector_resampled: shape portafogli*scenari #returns_vector_resampled: shape portafogli*scenari #shape del risultato: portafogli*scenari*tipo_variabile (dove tipo_variabile é pari a 2, ovvero: deviazioni standard e rendimenti) array_resampled_frontiers=np.stack([std_dev_annualized_vector_resampled*100, returns_vector_resampled*100], axis=2) #creiamo e visualizziamo il grafico inerente le singole frontiere di Markowitz ricampionate fig = go.Figure([ go.Scatter( x=array_resampled_frontiers[:, i, 0], y=array_resampled_frontiers[:, i, 1], mode='lines', #marker=dict(color="#444"), line=dict(width=1), showlegend=False, hoverinfo='none' ) for i in range(n_scenari)]) fig.update_layout( yaxis_title='Return', xaxis_title='Standard Deviation', title='Resampled Frontiers', xaxis={"tickformat": ".2f"}, yaxis={"tickformat": ".2f"}, width=1500, height=1000 ) if n_scenari<=1000 or forza_plotting_interattivo: st.plotly_chart(fig, staticPlot=True, use_container_width=True) #possibile aggiungere render_mode='auto' (oppure "webgl" o "svg") else: fig_png=pio.to_image(fig.update_layout(template='plotly_dark'), format="png", width=1500, height=1000, scale=None, validate=True, engine='auto') st.image(fig_png) #creiamo e visualizziamo il grafico inerente le singole frontiere di Markowitz ricampionate #rappresentate sotto forma di punti invece che linee continue fig = go.Figure([ go.Scatter( x=array_resampled_frontiers[:, i, 0], y=array_resampled_frontiers[:, i, 1], mode='markers', marker=dict(color='LightSkyBlue', size=3), opacity=0.5, #effetto gradevole ma peggiora le performance #line=dict(width=1), showlegend=False, hoverinfo='none' ) for i in range(n_scenari)]) fig.update_layout( yaxis_title='Return', xaxis_title='Standard Deviation', title='Resampled Frontiers (singoli portafogli)', xaxis={"tickformat": ".2f"}, yaxis={"tickformat": ".2f"}, width=1500, height=1000 ) if n_scenari<=1000 or forza_plotting_interattivo: st.plotly_chart(fig, staticPlot=True, use_container_width=True) #possibile aggiungere render_mode='auto' (oppure "webgl" o "svg") else: fig_png=pio.to_image(fig.update_layout(template='plotly_dark'), format="png", width=1500, height=1000, scale=None, validate=True, engine='auto') st.image(fig_png) #plottiamo la frontiera di Markowitz vs la Frontiera Ricampionata nel caso i rendimenti attesi e le volatilitá e correlazioni #siano perfette (nel qual caso ovviamente Markowitz si trova necessariamente sopra la Frontiera Ricampionata) fig = go.Figure([ go.Scatter( name='Frontiera di Markowitz', x=std_dev_annualized_vector*100, y=returns_vector*100, mode='lines', line=dict(width=1), showlegend=True ), go.Scatter( name='Frontiera Ricampionata', x=dev_std_frontiera_ricampionata*100, y=rendimenti_frontiera_ricampionata*100, mode='lines', line=dict(width=1), showlegend=True, hoverinfo='none' ) ]) fig.update_layout( yaxis_title='Return', xaxis_title='Standard Deviation', title='Resampled Frontiers vs Markowitz', xaxis={"tickformat": ".2f"}, yaxis={"tickformat": ".2f"}, width=1500, height=1000 ) st.plotly_chart(fig, staticPlot=True, use_container_width=True) #di seguito si ricalcolano rendimenti e rischi di tutti i portafogli simulati #ipotizzando che lo scenario base sia corretto (dunque usando i suoi rendimenti #attesi e la sua matrice di covarianza) if plotting_grafico_ptf_simulati_spazio_originario=="y": #inizializza le matrici destinate a contenere varianze e ritorni (da resampling) delle varie frontiere efficienti ricalcolate sullo scenario base (rendimenti attesi e matrice covarianza originali) variance_vector_2darray_recalc=np.zeros([(n_step+1), n_scenari]) returns_vector_2darray_recalc=np.zeros([(n_step+1), n_scenari]) #effettua il ricalcolo per singolo portafoglio for s in range(n_scenari): for p in range(n_step+1): variance_vector_2darray_recalc[p, s]=fun_varianza(ptf_weights_3darray[:, p, s]) returns_vector_2darray_recalc[p, s]=-fun_rendimento(ptf_weights_3darray[:, p, s]) #si usa il segno meno perchè la funzione restituiva l'opposto causa uso con il minimizer #calcola le deviazioni standard annualizzate std_dev_ann_vector_2darray_recalc=np.sqrt(variance_vector_2darray_recalc)*np.sqrt(periodicity) ##shape degli array numpy coinvolti di seguito #array_resampled_frontiers_spazio_originario=np.stack([(std_dev_ann_vector_2darray_recalc.flatten())*100, (returns_vector_2darray_recalc.flatten())*100], axis=2) #creiamo e visualizziamo il grafico inerente le singole frontiere di Markowitz ricampionate fig = go.Figure([ go.Scatter( x=(std_dev_ann_vector_2darray_recalc.flatten())*100, y=(returns_vector_2darray_recalc.flatten())*100, mode='markers', marker=dict(color='LightSkyBlue', size=3), opacity=0.5, #effetto gradevole ma peggiora le performance showlegend=False, hoverinfo='none' )]) fig.update_layout( yaxis_title='Return', xaxis_title='Standard Deviation', title="Portafogli simulati in spazio rischio-rendimento originario", xaxis={"tickformat": ".2f"}, yaxis={"tickformat": ".2f"}, width=1500, height=1000 ) if n_scenari<=1000 or forza_plotting_interattivo: st.plotly_chart(fig, staticPlot=True, use_container_width=True) #possibile aggiungere render_mode='auto' (oppure "webgl" o "svg") else: fig_png=pio.to_image(fig.update_layout(template='plotly_dark'), format="png", width=1500, height=1000, scale=None, validate=True, engine='auto') st.image(fig_png) #si recuperano e si plottano gli andamenti di portafoglio nei diversi portafogli/scenari if plottare_andamenti_simulati=="y": for n_ro_ptf in range(0, n_step+1, step_ptf_resamp_graph): #inizializza la matrice destinata a contenere i rendimenti di portafoglio relativi ai diversi scenari portfolio_returns_scenari=np.zeros([n, n_scenari]) #qui si calcolano i rendimenti di portafoglio partendo dai rendimenti ricampionati degli asset ed i relativi pesi (da frontiera ricampionata) for z in range(n_scenari): portfolio_returns_scenari[:, z]=(output_t_3d[:, :, z]).dot(ptf_weights_resampled[:, n_ro_ptf]) #di seguito si setta il numero di periodi di tempo (settimane o mesi) in modo tale che l'orizzonte temporale del grafico non superi l'anno tempo_grafico=min(n, periodicity) #inizializza la matrice destinata ad accogliere i valori del portafoglio nel suo andamento temporale e per i diversi scenari andamento_portafoglio=np.zeros([tempo_grafico+1, n_scenari]) #ovviamente si parte dal valore di 100 ad inizio periodo andamento_portafoglio[0, :]=np.array([100 for i in range(n_scenari)]) #di seguito si calcola l'andamento del portafoglio nel tempo e per i diversi scenari for z in range(n_scenari): for t in range(tempo_grafico): andamento_portafoglio[t+1, z]=andamento_portafoglio[t, z]*(1+portfolio_returns_scenari[t, z]) #plottiamo gli andamenti di portafoglio fig = go.Figure([ go.Scatter( #name='Resampled Frontier', #x=andamento_portafoglio[:, i], y=andamento_portafoglio[:, i], mode='lines', line=dict(width=0.5, color='#c44e52'), #'#c44e52' oppure '#d62728' showlegend=False, hoverinfo='none' ) for i in range(n_scenari)]) fig.update_layout( yaxis_title="Valore del portafoglio (fatto 100 iniziale)", xaxis_title="Tempo", title=f"Portafoglio simulato n.ro {n_ro_ptf} (da Resampled Fr.)", width=1500, height=1000 ) st.plotly_chart(fig, staticPlot=True, use_container_width=True) ``` ### Steps To Reproduce please read only the error ### Expected Behavior cache the function ### Current Behavior AttributeError Traceback: File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py", line 565, in _run_script exec(code, module.__dict__) File "C:\Users\A260219\Jupyter\bloomberg\prova.py", line 2126, in <module> ottimizzazione(num_variabili, fun_varianza, ptf_weights_equally, expected_returns, lista_asset_pairplot) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cache_utils.py", line 186, in __call__ return self._get_or_create_cached_value(args, kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cache_utils.py", line 211, in _get_or_create_cached_value return self._handle_cache_miss(cache, value_key, func_args, func_kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cache_utils.py", line 265, in _handle_cache_miss computed_value = self._info.func(*func_args, **func_kwargs) File "C:\Users\A260219\Jupyter\bloomberg\prova.py", line 1555, in ottimizzazione st.write(pd.DataFrame(data=np.round(ptf_weights.reshape([1, num_variabili])*100, decimals=1), columns=dati_ritorni.columns, index=["Pesi"])) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\metrics_util.py", line 311, in wrapped_func result = non_optional_func(*args, **kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\elements\write.py", line 200, in write self.dg.dataframe(arg) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\metrics_util.py", line 311, in wrapped_func result = non_optional_func(*args, **kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\elements\dataframe_selector.py", line 116, in dataframe return self.dg._arrow_dataframe( File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\metrics_util.py", line 311, in wrapped_func result = non_optional_func(*args, **kwargs) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\elements\arrow.py", line 117, in _arrow_dataframe return self.dg._enqueue("arrow_data_frame", proto) File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\delta_generator.py", line 563, in _enqueue caching.save_element_message( File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\__init__.py", line 50, in save_element_message CACHE_RESOURCE_MESSAGE_REPLAY_CTX.save_element_message( File "C:\Users\A260219\Anaconda3\envs\lavoro\lib\site-packages\streamlit\runtime\caching\cached_message_replay.py", line 293, in save_element_message raise AttributeError ### Is this a regression? - [X] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.18 - Python version: 3.10 - Operating System: windows - Browser: chrome - Virtual environment: - ### Additional Information no problems at all with st.experimental_singleton (streamlit version 1.17) but with st.cache_resource in streamlit 1.18 i get the error above the function is very very large and computationally expensive, but the behavior was very fine with st.experimental_singleton and streamlit version 1.17 ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6103/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6102

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6102/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6102/comments

https://api.github.com/repos/streamlit/streamlit/issues/6102/events

https://github.com/streamlit/streamlit/issues/6102

I_kwDODCoeTs5eObuS

Add logo extra not working

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary add logo extra is no longer working either in my project or to visualise on streamlit extras website ### Reproducible Code Example ```Python import streamlit as st from streamlit_extras.app_logo import add_logo add_logo("https://logos-world.net/wp-content/uploads/2020/06/Manchester-United-logo.png") ``` ### Steps To Reproduce _No response_ ### Expected Behavior _No response_ ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: - Python version: - Operating System: - Browser: - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6102/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6101

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6101/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6101/comments

https://api.github.com/repos/streamlit/streamlit/issues/6101/events

https://github.com/streamlit/streamlit/pull/6101

PR_kwDODCoeTs5JwDQ-

Up version to 1.18.1

open

COLLABORATOR

https://api.github.com/repos/streamlit/streamlit/issues/6101/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6100

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6100/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6100/comments

https://api.github.com/repos/streamlit/streamlit/issues/6100/events

https://github.com/streamlit/streamlit/pull/6100

PR_kwDODCoeTs5Ju4Y_

Check availability of local storage to prevent exceptions

closed

COLLABORATOR

## 📚 Context As stated in https://github.com/streamlit/streamlit/issues/6087, iframe embedded Streamlit apps might break in Chrome wit third party cookie blocking setting. The main problem is that we are accessing the localStorage for `sidebarWidth` without properly checking the availability of localStorage. - What kind of change does this PR introduce? - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - Move `localStorageAvailable` method to `storageUtils` - Check availability for localStorage before caching the current sidebar width (the caching here is more of an optional feature anyways). ## 🧪 Testing Done - [ ] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: Closes #6087 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6100/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6099

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6099/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6099/comments

https://api.github.com/repos/streamlit/streamlit/issues/6099/events

https://github.com/streamlit/streamlit/issues/6099

I_kwDODCoeTs5eLacP

[Altair] `st.altair_chart` removes data from a compound plot with interaction that works with `altair`

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary Hello there, When doing a compound plot in altair with interval interactions, the lines are not displayed properly. The images below will make it very clear. Thanks a lot for your help ! ### Reproducible Code Example ```python # requirements.txt streamlit==1.18.0 altair==4.2.2 pandas==1.3.3 altair_viewer==0.4.0 ``` ```python # webapp.py import streamlit as st from altair_demo import main_altair def main(): st.write("Hello") chart = main_altair() st.altair_chart(altair_chart=chart) if __name__ =="__main__": main() ``` ```python # altair_demo.py import pandas as pd import altair as alt alt.renderers.enable('mimetype') def main(): data = pd._testing.makeTimeDataFrame(freq='H') data = data.reset_index().melt(id_vars="index") interval = alt.selection_interval(encodings=["x"]) base = ( alt.Chart(data=data) .mark_line(point={"filled": False, "fill": "white"}, strokeWidth=2) .encode( x=alt.X("index:T", title="Date"), y=alt.Y("value", title="Cible"), ) ) line_chart = ( base.encode( x=alt.X("index:T", title="Date", scale=alt.Scale(domain=interval.ref())), color="variable", ) ).properties(width=600, height=300) view = ( base.add_selection(interval) .encode(color="variable") .properties( width=600, height=50, ) ) histogram = ( alt.Chart(data=data) .mark_bar().encode( alt.X("value:Q", bin=True), y=alt.Y("count()"), ) ).transform_filter( interval ) return line_chart & view | histogram if __name__ =="__main__": chart = main() # chart.show() st.altair_chart(chart) ``` ### Steps To Reproduce 0. Create the files `requirements.txt`, `webapp.py` and `altair_demo.py` 1. `pip install -r requirements.txt` 2. `python altair_demo.py` -> Displaying chart at http://localhost:45747/ 3. `streamlit run webapp.py` -> http://localhost:8501 ### Expected Behavior Pictures speak volume. Altair plot from the code below:  ### Current Behavior Streamlit equivalent of the same plot:  ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.18.0 - Python version: Python 3.8.10 - Operating System: Ubuntu 20.04 on Windows 11 WSL - Browser: Firefox 109.0.1 (64-bit) - Virtual environment: virtualenv ### Additional Information I think the issue was already raised here first: https://discuss.streamlit.io/t/data-missing-from-altair-compound-chart/11527 ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6099/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6098

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6098/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6098/comments

https://api.github.com/repos/streamlit/streamlit/issues/6098/events

https://github.com/streamlit/streamlit/pull/6098

PR_kwDODCoeTs5Jum0s

Improve readability of styled cells in dark mode

closed

COLLABORATOR

## 📚 Context As mentioned in this PR, the default usage of Pandas Styler's `highlight_max` within an app used in dark mode will result in unreadable cells (https://github.com/streamlit/streamlit/issues/6060). This PR adds a quick fix that uses our dark font color in case the cell background is set to yellow. Yellow is used as default in Pandas Styler in case the user does not specify any specific color. Thereby, the default `highlight_max` use works out of the box with any Streamlit theme. - What kind of change does this PR introduce? - [ ] Bugfix - [x] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - Use a dark color in case the background is `yellow` (default in `highlight_max`). **Revised:**  **Current:**  ## 🧪 Testing Done - [x] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: Closes #6060 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6098/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6097

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6097/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6097/comments

https://api.github.com/repos/streamlit/streamlit/issues/6097/events

https://github.com/streamlit/streamlit/issues/6097

I_kwDODCoeTs5eJ5cz

Connecting to an API with win32com results in com_error after 30secs

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary I am trying to build an app that will connect into the API of another running app on Windows through the com interface. This has been successfully done using Jupyter and win32com. No issues have been experienced connecting in and re-running the cell multiple times. However, when attempting the same in Streamlit, I am able to get it working for about 30 secs. After that and when the app is updated or rerun it results in a com_error. I have recreated the issue using Excel and the same behavior is experienced, which rules out the application I am trying to build this for. The code I am using is an adapted version of this SO post: [Python Multithreading COM Objects - Stack Overflow](https://stackoverflow.com/questions/51670530/python-multithreading-com-objects) ### Reproducible Code Example ```Python import streamlit as st import win32com.client import pythoncom st.header('Test') com = win32com.client.Dispatch("Excel.Application") com_id = win32com.client.pythoncom.CoMarshalInterThreadInterfaceInStream( win32com.client.pythoncom.IID_IDispatch, com) pythoncom.CoInitialize() api = win32com.client.Dispatch(pythoncom.CoGetInterfaceAndReleaseStream(com_id, pythoncom.IID_IDispatch)) add_wb = st.button('Add Workbook') if add_wb : api.Workbooks.Add() ``` ### Steps To Reproduce 1. Run the above Streamlit App 2. Open Excel 3. Click the Add Workbook button and Excel will (should) create new instances/workbooks each time it is pressed 4. Try again after 30-60s and the com_error will appear ### Expected Behavior No com_error to occur. ### Current Behavior After 30-60s the following error appears: com_error: (-2147221008, ‘CoInitialize has not been called.’, None, None) File "C:\Users\Andy\AppData\Local\Programs\Python\Python39\lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py", line 562, in _run_script exec(code, module.__dict__) File "C:\PythonProjects\streamlit_integration\test_app.py", line 7, in <module> com = win32com.client.Dispatch("Excel.Application") File "C:\Users\Andy\AppData\Roaming\Python\Python39\site-packages\win32com\client\__init__.py", line 117, in Dispatch dispatch, userName = dynamic._GetGoodDispatchAndUserName(dispatch, userName, clsctx) File "C:\Users\Andy\AppData\Roaming\Python\Python39\site-packages\win32com\client\dynamic.py", line 106, in _GetGoodDispatchAndUserName return (_GetGoodDispatch(IDispatch, clsctx), userName) File "C:\Users\Andy\AppData\Roaming\Python\Python39\site-packages\win32com\client\dynamic.py", line 88, in _GetGoodDispatch IDispatch = pythoncom.CoCreateInstance( ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.13.0 - Python version: 3.9.13 64 Bit - Operating System: Win 10 - Browser: Edge 108.0.1462.76 - Virtual environment: None ### Additional Information Upgraded Streamlit to 1.18.0. The above code no longer generates a new Excel workbook, but the com_error still occurs after 30-60secs ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6097/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6096

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6096/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6096/comments

https://api.github.com/repos/streamlit/streamlit/issues/6096/events

https://github.com/streamlit/streamlit/issues/6096

I_kwDODCoeTs5eJtXg

Streamlit theme not working on Safari and Google Chrome

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary St shows automatically a custom theme I dunno where to modify when opening on Safari, Google Chrome and Firefox. They don't follow the theme I made in .streamlit/config.toml . Is there a way to set the global theme? ### Reproducible Code Example ```Python This is the theme it shows [theme] base="dark" primaryColor="#29a2bb" backgroundColor="#afc6d2" secondaryBackgroundColor="#90d4de" ``` ### Steps To Reproduce _No response_ ### Expected Behavior I tried to put the light theme in my config.toml, but it still shows the old theme in custom theme. ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version:1.13.0 - Python version:3.82 - Operating System:Mac OS - Browser:Safari, Google Chrome, Firefox - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6096/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6095

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6095/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6095/comments

https://api.github.com/repos/streamlit/streamlit/issues/6095/events

https://github.com/streamlit/streamlit/issues/6095

I_kwDODCoeTs5eJDlw

Rename endpoints #5534 breaks streamlit behind load balancer

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary Hi Team, with the changes introduced in Rename endpoints #5534 you changed the websocket request path from `/stream` to `/_stcore/stream`. We are running streamlit behind a load balancer with a specific configuration to upgrade requests that reach `/stream` to use Websocket. This is broken since the latest release and our Apps with the new streamlit version are not working anymore. I am pretty sure, we are not the only organization running into this issue. If it's possible, could you rollback this change? Thanks, Nicolas ### Reproducible Code Example _No response_ ### Steps To Reproduce _No response_ ### Expected Behavior _No response_ ### Current Behavior _No response_ ### Is this a regression? - [X] Yes, this used to work in a previous version. ### Debug info - Streamlit version: streamlit-1.18.0 - Python version: python-3.8.15 - Operating System: macOS 13.1 - Browser: Version 109.0.5414.119 (Official Build) (x86_64) - Virtual environment: ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6095/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6094

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6094/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6094/comments

https://api.github.com/repos/streamlit/streamlit/issues/6094/events

https://github.com/streamlit/streamlit/pull/6094

PR_kwDODCoeTs5JsgU5

Downgrade decamelize to avoid Safari break

closed

COLLABORATOR

## 📚 Context Streamlit 1.18.0 breaks on Safari due to [lack of support of Regex lookbehinds](https://caniuse.com/js-regexp-lookbehind) The best solution for now is to downgrade what caused the issue. - What kind of change does this PR introduce? - [x] Bugfix ## 🧠 Description of Changes Downgrade `decamelize` to avoid lookbehinds. Safari does not support this feature. Downgrading required a downgrade in `humanize-string` This change reverts the specific piece from #5946 if you are curious that the revert returns these dependencies back to normal. ## 🧪 Testing Done We cannot test specifically on Safari, which stinks a lot, so manual testing and our current CI tests should be sufficient. ## 🌐 References - **Issue**: Closes #6092 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6094/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6093

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6093/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6093/comments

https://api.github.com/repos/streamlit/streamlit/issues/6093/events

https://github.com/streamlit/streamlit/issues/6093

I_kwDODCoeTs5eIb6v

Regression in v1.18.0 `st.metric`'s `help` tooltip position

open

MEMBER

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary [Reported](https://discuss.streamlit.io/t/version-1-18-0/37501/3?u=snehankekre) by @marduk2 on our forum. The position of the `help` tooltip for `st.metric` has changed positions in v1.18.0 and appears under the metric label instead of to its right. ### Reproducible Code Example ```Python import streamlit as st st.title("'Help' tooltips in st.metric render in a weird position") st.metric( label="Example metric", help="Something doesn't feel right...", value=150.59, delta="Very high", ) ``` ### Steps To Reproduce 1. Run the app ### Expected Behavior  ### Current Behavior  ### Is this a regression? - [X] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.18.0 - Python version: 3.9 - Operating System: macOS - Browser: Chrome - Virtual environment: NA ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6093/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6092

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6092/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6092/comments

https://api.github.com/repos/streamlit/streamlit/issues/6092/events

https://github.com/streamlit/streamlit/issues/6092

I_kwDODCoeTs5eHwfj

Streamlit 1.18.0 and Safari 16.3 invalid regular expression error

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary A simple st.write('Hello world') app fails in Safari 16.3 using streamlit 1.18.0 with this error [Error] SyntaxError: Invalid regular expression: invalid group specifier name (anonymous function) (main.223e9539.js:2) The app works in Safari 16.3 with streamlit 1.17.0. The app also works in Firefox 109.0.1 and Chrome 110.0.5481.77 with streamlit 1.17.0 and 1.18.0 ### Reproducible Code Example ```Python import streamlit as st st.write(''' # My first app Hello *world!* ''') ``` ### Steps To Reproduce streamlit run sthelloworld.py ### Expected Behavior Should see Hello world in a new Safari window ### Current Behavior See a blank screen in Safari. The console shows the error [Error] SyntaxError: Invalid regular expression: invalid group specifier name (anonymous function) (main.223e9539.js:2) ### Is this a regression? - [X] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.18.0 - Python version: 3.11.2 and 3.10.10 - Operating System: macOS 13.2 - Browser: Safari 16.3 (18614.4.6.1.5) - Virtual environment: vent ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6092/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6091

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6091/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6091/comments

https://api.github.com/repos/streamlit/streamlit/issues/6091/events

https://github.com/streamlit/streamlit/issues/6091

I_kwDODCoeTs5eHBDr

chatgpt

closed

NONE

_(Note, you don't have to fill out every section here. They're just here for guidance. That said, nicely detailed feature requests are more likely to get eng attention sooner)_ ### Problem Is your feature request related to a problem? Please describe the problem here. Ex. I'm always frustrated when [...] ### Solution **MVP:** What's the smallest possible solution that would get 80% of the problem out of the way? **Possible additions:** What are other things that could be added to the MVP over time to make it better? **Preferred solution:** If you don't like the MVP above, tell us why, and what you'd like done instead. ### Additional context Add any other context or screenshots about the feature request here. For example, did this FR come from https://discuss.streamlit.io or another site? Link the original source here! --- Community voting on feature requests enables the Streamlit team to understand which features are most important to our users. **If you'd like the Streamlit team to prioritize this feature request, please use the 👍 (thumbs up emoji) reaction in response to the initial post.**

https://api.github.com/repos/streamlit/streamlit/issues/6091/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6090

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6090/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6090/comments

https://api.github.com/repos/streamlit/streamlit/issues/6090/events

https://github.com/streamlit/streamlit/pull/6090

PR_kwDODCoeTs5Jp53u

up version to 1.18.0

closed

COLLABORATOR

https://api.github.com/repos/streamlit/streamlit/issues/6090/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6089

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6089/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6089/comments

https://api.github.com/repos/streamlit/streamlit/issues/6089/events

https://github.com/streamlit/streamlit/issues/6089

I_kwDODCoeTs5eFiie

Add datetime input

open

NONE

### Problem The st.date_input() is not taken into account the time in the widget. We can only select : day, month, year. In order to bypass this limitation, I’m using st.text_input, next I convert it into full datetime format (I don’t want to use st.slider(), It’ s ugly).  ### Solution The purpose is to add time selection in the widget like this :  -------------------------------------------- Community voting on feature requests enables the Streamlit team to understand which features are most important to our users. If you'd like the Streamlit team to prioritize this feature request, please use the 👍 (thumbs up emoji) reaction in response to the initial post.

https://api.github.com/repos/streamlit/streamlit/issues/6089/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6088

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6088/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6088/comments

https://api.github.com/repos/streamlit/streamlit/issues/6088/events

https://github.com/streamlit/streamlit/pull/6088

PR_kwDODCoeTs5Jpi47

Avoid showing emoji on non-UTF-8 terminals.

open

NONE

Without this, new users get cryptic encoding errors. Fixes: https://github.com/streamlit/streamlit/issues/2284 ## 📚 Context - What kind of change does this PR introduce? - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - Added a condition to only show emoji when the encoding is utf-8. - There may be other terminal encodings that also work, but utf-8 is the de facto standard. - It's unclear whether the remaining condition about Windows terminals is needed; it's not easy for me to test that. So I've left in that condition to be conservative. - [ ] This is a breaking API change - [ ] This is a visible (user-facing) change ## 🧪 Testing Done To test: 1. Remove streamlit config (so it prints the greeting) 2. Run `env PYTHONIOENCODING=latin-1 streamlit hello` 3. It shouldn't crash (but it won't print the emoji greeting). --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6088/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6087

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6087/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6087/comments

https://api.github.com/repos/streamlit/streamlit/issues/6087/events

https://github.com/streamlit/streamlit/issues/6087

I_kwDODCoeTs5eCWdy

iframe-embedded Streamlit apps do not work on recent version Chrome with blocking third party cookies setting

closed

CONTRIBUTOR

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary Streamlit apps that are embedded through iframe do not appear when we use Chrome (or Chrome-based browsers) with third party cookie blocking setting. For example, see the sample embedded in the official document below. https://docs.streamlit.io/library/api-reference/write-magic/st.write Without the cookie-blocking setting, it works as below. <img width="1480" alt="CleanShot 2023-02-09 at 19 52 08@2x" src="https://user-images.githubusercontent.com/3135397/217792545-17dd1fd0-acaa-425f-9e9a-897b0d5d405b.png"> However, if we enable the settings as below, <img width="765" alt="CleanShot 2023-02-09 at 19 53 08@2x" src="https://user-images.githubusercontent.com/3135397/217792764-03020f74-1129-47e7-a2cc-394de0d27dea.png"> the embedded app does not appear.  FYI: According to [this Chromium document](https://www.chromium.org/for-testers/bug-reporting-guidelines/uncaught-securityerror-failed-to-read-the-localstorage-property-from-window-access-is-denied-for-this-document/), it might be an intended security behavior. ### Reproducible Code Example _No response_ ### Steps To Reproduce _No response_ ### Expected Behavior Embedding apps through iframe is officially expected usage, so it should work even in this setting (maybe ignoring the errors at localStorage access and fallback to some default values?) ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: (Official sample on the Cloud) - Python version: (Official sample on the Cloud) - Operating System: macOS Ventura 13.1 (22C65) - Browser: Version: Chrome Version 110.0.5481.77 (Official Build) (arm64), SideKick Version 108.36.1.29778 (Official Build) (arm64) - Virtual environment: (Official sample on the Cloud) ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6087/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6086

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6086/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6086/comments

https://api.github.com/repos/streamlit/streamlit/issues/6086/events

https://github.com/streamlit/streamlit/issues/6086

I_kwDODCoeTs5d_bPW

No border around `st.form`

open

COLLABORATOR

### Problem Sometimes the border around forms doesn't look good visually. ### Solution Add a parameter to remove it. --- Community voting on feature requests enables the Streamlit team to understand which features are most important to our users. **If you'd like the Streamlit team to prioritize this feature request, please use the 👍 (thumbs up emoji) reaction in response to the initial post.**

https://api.github.com/repos/streamlit/streamlit/issues/6086/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6085

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6085/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6085/comments

https://api.github.com/repos/streamlit/streamlit/issues/6085/events

https://github.com/streamlit/streamlit/issues/6085

I_kwDODCoeTs5d_aFy

Markdown support for radio buttons

open

COLLABORATOR

### Problem Colored text and other markdown elements work in the label of `st.radio` but not in the texts of the radio elements. This is a bit weird since we do support it in the texts of checkboxes (where the text next to the checkbox is the label). ### Solution Allow markdown in the options of `st.radio`. --- Community voting on feature requests enables the Streamlit team to understand which features are most important to our users. **If you'd like the Streamlit team to prioritize this feature request, please use the 👍 (thumbs up emoji) reaction in response to the initial post.**

https://api.github.com/repos/streamlit/streamlit/issues/6085/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6084

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6084/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6084/comments

https://api.github.com/repos/streamlit/streamlit/issues/6084/events

https://github.com/streamlit/streamlit/pull/6084

PR_kwDODCoeTs5JkCo_

Refactor Radio components

open

COLLABORATOR

## 📚 Context Tech Debt: Refactor `Radio.tsx` in `shared/Radio` & `widgets/Radio` - What kind of change does this PR introduce? - [x] Refactoring ## 🧠 Description of Changes - Refactored both `Radio.tsx` files into functional components w/ hooks - Refactored both `Radio.test.tsx` files to use `react-testing-library` instead of enzyme ## 🧪 Testing Done - [x] Added/Updated unit tests

https://api.github.com/repos/streamlit/streamlit/issues/6084/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6083

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6083/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6083/comments

https://api.github.com/repos/streamlit/streamlit/issues/6083/events

https://github.com/streamlit/streamlit/pull/6083

PR_kwDODCoeTs5JjdhZ

Allow full screen button to show for dataframes, charts, etc in expander

closed

COLLABORATOR

<!-- Before contributing (PLEASE READ!) ⚠️ If your contribution is more than a few lines of code, then prior to starting to code on it please post in the issue saying you want to volunteer, then wait for a positive response. And if there is no issue for it yet, create it first. This helps make sure: 1. Two people aren't working on the same thing 2. This is something Streamlit's maintainers believe should be implemented/fixed 3. Any API, UI, or deeper architectural changes that need to be implemented have been fully thought through by Streamlit's maintainers 4. Your time is well spent! More information in our wiki: https://github.com/streamlit/streamlit/wiki/Contributing --> ## 📚 Context https://github.com/streamlit/streamlit/issues/5694 _Please describe the project or issue background here_ Fullscreen button can't be found because the overflow is hidden so just setting the parent div to overflow: visible will solve this. Solution was found by: @LukasMasuch - What kind of change does this PR introduce? - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - _Add bullet points summarizing your changes here_ - [ ] This is a breaking API change - [x] This is a visible (user-facing) change **Revised:** _Insert screenshot of your updated UI/code here_  **Current:**  _Insert screenshot of existing UI/code here_ ## 🧪 Testing Done - [x] Screenshots included - [ ] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: Closes #5694 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6083/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6082

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6082/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6082/comments

https://api.github.com/repos/streamlit/streamlit/issues/6082/events

https://github.com/streamlit/streamlit/pull/6082

PR_kwDODCoeTs5JjTzN

Fix StreamlitDialog "Clear Cache" description

closed

COLLABORATOR

The "Clear Cache" confirm dialog now refers to "cache_data/resource" instead of "experimental_memo/singleton"

https://api.github.com/repos/streamlit/streamlit/issues/6082/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6081

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6081/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6081/comments

https://api.github.com/repos/streamlit/streamlit/issues/6081/events

https://github.com/streamlit/streamlit/issues/6081

I_kwDODCoeTs5d93jq

x-axis display error st.plotly_chart

closed

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary Incorrect display of the graph streamlit (figure 1). A similar graph is displayed correctly in jupiter notebook figure (2). Code to reproduce the problem.   ### Reproducible Code Example ```Python #jupyter import pandas as pd import plotly.express as px df1=pd.DataFrame([[403, 'April', 686], [403, 'August', 958], [403, 'December', 653], [403, 'February', 645], [403, 'January', 552], [403, 'July', 901], [403, 'June', 818], [403, 'March', 587], [403, 'May', 719], [403, 'November', 850], [403, 'October', 851], [403, 'September', 933], [601, 'April', 28], [601, 'August', 35], [601, 'December', 14], [601, 'February', 13], [601, 'January', 17], [601, 'July', 32], [601, 'June', 27], [601, 'March', 18], [601, 'May', 15], [601, 'November', 14], [601, 'October', 24], [601, 'September', 21], [602, 'April', 49], [602, 'August', 169], [602, 'December', 123], [602, 'February', 64], [602, 'January', 64], [602, 'July', 132], [602, 'June', 146], [602, 'March', 77], [602, 'May', 65], [602, 'November', 106], [602, 'October', 225], [602, 'September', 159], [603, 'April', 10], [603, 'August', 5], [603, 'December', 8], [603, 'February', 18], [603, 'January', 15], [603, 'July', 5], [603, 'June', 2], [603, 'March', 7], [603, 'May', 18], [603, 'November', 14], [603, 'October', 18], [603, 'September', 6], [701, 'April', 5], [701, 'August', 7], [701, 'December', 2], [701, 'February', 1], [701, 'January', 2], [701, 'July', 6], [701, 'June', 7], [701, 'March', 5], [701, 'May', 10], [701, 'November', 2], [701, 'October', 2], [701, 'September', 3], [801, 'April', 1], [801, 'August', 12], [801, 'December', 16], [801, 'February', 1], [801, 'January', 1], [801, 'July', 9], [801, 'June', 5], [801, 'March', 1], [801, 'May', 1], [801, 'November', 16], [801, 'October', 16], [801, 'September', 16], ['H99', 'April', 18], ['H99', 'August', 3], ['H99', 'December', 1], ['H99', 'February', 13], ['H99', 'January', 12], ['H99', 'July', 7], ['H99', 'June', 2], ['H99', 'March', 17], ['H99', 'May', 21], ['H99', 'November', 5], ['H99', 'October', 3], ['H99', 'September', 2], ['H99', 'December', 2], ['H99', 'May', 2], ['H99', 'November', 2]], columns=['PP', 'Month', 'Count']) df1["PP"] = df1["PP"].astype(str) months=['January', 'February', 'March', 'April', 'May','June', 'July', 'August', 'September', 'October', 'November', 'December'] PP=["H99", "403", "601", "602", "603", "701", "801"] fig1 = px.histogram( df1, x="PP", y="Count", color="Month", barmode="group", height=550, category_orders={ "PP": PP, "Month": months, }, text_auto=True, ) fig1.update_layout(title_text="P", title_x=0.5) fig1.update_xaxes( title_text="<b>PP</b>", ) fig1.update_yaxes( title_text="<b>Count</b>", secondary_y=False, ) fig1.update_layout(hovermode="x unified") fig1.update_traces( hovertemplate="<b>Count</b>: %{y:r}", textposition="auto", textangle=0, ) fig1.show() #streamlit import pandas as pd import plotly.express as px import streamlit as st df1=pd.DataFrame([[403, 'April', 686], [403, 'August', 958], [403, 'December', 653], [403, 'February', 645], [403, 'January', 552], [403, 'July', 901], [403, 'June', 818], [403, 'March', 587], [403, 'May', 719], [403, 'November', 850], [403, 'October', 851], [403, 'September', 933], [601, 'April', 28], [601, 'August', 35], [601, 'December', 14], [601, 'February', 13], [601, 'January', 17], [601, 'July', 32], [601, 'June', 27], [601, 'March', 18], [601, 'May', 15], [601, 'November', 14], [601, 'October', 24], [601, 'September', 21], [602, 'April', 49], [602, 'August', 169], [602, 'December', 123], [602, 'February', 64], [602, 'January', 64], [602, 'July', 132], [602, 'June', 146], [602, 'March', 77], [602, 'May', 65], [602, 'November', 106], [602, 'October', 225], [602, 'September', 159], [603, 'April', 10], [603, 'August', 5], [603, 'December', 8], [603, 'February', 18], [603, 'January', 15], [603, 'July', 5], [603, 'June', 2], [603, 'March', 7], [603, 'May', 18], [603, 'November', 14], [603, 'October', 18], [603, 'September', 6], [701, 'April', 5], [701, 'August', 7], [701, 'December', 2], [701, 'February', 1], [701, 'January', 2], [701, 'July', 6], [701, 'June', 7], [701, 'March', 5], [701, 'May', 10], [701, 'November', 2], [701, 'October', 2], [701, 'September', 3], [801, 'April', 1], [801, 'August', 12], [801, 'December', 16], [801, 'February', 1], [801, 'January', 1], [801, 'July', 9], [801, 'June', 5], [801, 'March', 1], [801, 'May', 1], [801, 'November', 16], [801, 'October', 16], [801, 'September', 16], ['H99', 'April', 18], ['H99', 'August', 3], ['H99', 'December', 1], ['H99', 'February', 13], ['H99', 'January', 12], ['H99', 'July', 7], ['H99', 'June', 2], ['H99', 'March', 17], ['H99', 'May', 21], ['H99', 'November', 5], ['H99', 'October', 3], ['H99', 'September', 2], ['H99', 'December', 2], ['H99', 'May', 2], ['H99', 'November', 2]], columns=['PP', 'Month', 'Count']) df1["PP"] = df1["PP"].astype(str) months=['January', 'February', 'March', 'April', 'May','June', 'July', 'August', 'September', 'October', 'November', 'December'] PP=["H99", "403", "601", "602", "603", "701", "801"] fig1 = px.histogram( df1, x="PP", y="Count", color="Month", barmode="group", height=550, category_orders={ "PP": PP, "Month": months, }, text_auto=True, ) fig1.update_layout(title_text="P", title_x=0.5) fig1.update_xaxes( title_text="<b>PP</b>", ) fig1.update_yaxes( title_text="<b>Count</b>", secondary_y=False, ) fig1.update_layout(hovermode="x unified") fig1.update_traces( hovertemplate="<b>Count</b>: %{y:r}", textposition="auto", textangle=0, ) st.plotly_chart(fig1, use_container_width=True, config={"displaylogo": False}) ``` ### Steps To Reproduce _No response_ ### Expected Behavior _No response_ ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.17.0 - Python version: 3.10.7 - Operating System: win 10 - Browser: google chrome 109.0 ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6081/timeline

completed

https://api.github.com/repos/streamlit/streamlit/issues/6080

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6080/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6080/comments

https://api.github.com/repos/streamlit/streamlit/issues/6080/events

https://github.com/streamlit/streamlit/pull/6080

PR_kwDODCoeTs5Ji0ws

Hamburger menu cleanup

open

NONE

…fies a custom URL. <!-- Before contributing (PLEASE READ!) ⚠️ If your contribution is more than a few lines of code, then prior to starting to code on it please post in the issue saying you want to volunteer, then wait for a positive response. And if there is no issue for it yet, create it first. This helps make sure: 1. Two people aren't working on the same thing 2. This is something Streamlit's maintainers believe should be implemented/fixed 3. Any API, UI, or deeper architectural changes that need to be implemented have been fully thought through by Streamlit's maintainers 4. Your time is well spent! More information in our wiki: https://github.com/streamlit/streamlit/wiki/Contributing --> ## 📚 Context _Please describe the project or issue background here_ This PR implements the hamburger menu items cleanup, according to the according to the "Clean up the hamburger menu" [Product spec](https://www.notion.so/streamlit/Product-spec-fa80e423ee5e4863963f5d0f4c873c7c). - What kind of change does this PR introduce? - [ ] Bugfix - [x] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - _Add bullet points summarizing your changes here_ * 🚧 Report a bug → Remove by default. BUT: We have a parameter `menu_items` in `st.set_page_config` that lets the developer manually set a link for this option (see [[here](https://docs.streamlit.io/library/api-reference/utilities/st.set_page_config)](https://docs.streamlit.io/library/api-reference/utilities/st.set_page_config)). If that is set (to something different than `None`), the item should still show up. (see below). No matter what, we should just use the name “Report a bug” going forward. * 🚧 Get help → Remove by default. BUT: We have a parameter `menu_items` in `st.set_page_config` that lets the developer manually set a link for this option (see [[here](https://docs.streamlit.io/library/api-reference/utilities/st.set_page_config)](https://docs.streamlit.io/library/api-reference/utilities/st.set_page_config)). If that is set (to something different than `None`), the item should still show up. * ❌ Streamlit Cloud → Remove. This is pretty redundant with “Deploy this app”. 0.4% of all clicks. * ❌ Report a Streamlit bug → Remove. Hardly gets clicks (0.0%). * ❌ Visit Streamlit docs → Remove. Hardly gets clicks (0.3%). * ❌ Visit Streamlit forums → Remove. Hardly gets clicks (0.1%). Changes: - [ ] This is a breaking API change - [x] This is a visible (user-facing) change **Revised:** _Insert screenshot of your updated UI/code here_ No custom menu items:  Deploy option available:  Custom menu items added using the following code ``` menu_items: MenuItems = { "Get help": "https://google.com", # must be a valid URL "Report a bug": "https://google.com", # must be a valid URL "About": "This is a custom about text set with st.set_page_config" # can be any text } st.set_page_config( page_title="This is custom page title", menu_items=menu_items ) ```  **Current:** _Insert screenshot of existing UI/code here_ <img width="222" alt="Screenshot 2023-02-07 at 10 31 20" src="https://user-images.githubusercontent.com/119880369/217660184-998ab1d5-ba4f-4c3c-89cb-c6d079d0ab89.png"> ## 🧪 Testing Done - [x] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: [STREAMLIT-2113](https://snowflakecomputing.atlassian.net/browse/STREAMLIT-2113) - **PR in s4t repo**: https://github.com/streamlit/s4t/pull/5087 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6080/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6079

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6079/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6079/comments

https://api.github.com/repos/streamlit/streamlit/issues/6079/events

https://github.com/streamlit/streamlit/issues/6079

I_kwDODCoeTs5d9vLX

session_state not actually changed until second page reload

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary **Summary:** When the multiselect widget is changed, the Streamlit page reloads from top to bottom, however the `session_state` receiving the output from the multiselect box [I.e. `session_state["myvar"] = st.multiselect(...)`] does not actually change (persistently) until the page reloads a _second_ time. **Details:** I don't know if this is actually a bug, a change request, or simply something that can't be done the way I'm coding it. When using a widget that passes it's results to session_state [I.e. `session_state["myvar"] = st.multiselect(...)`] **the page reloads from top to bottom _as soon_ as a selection is made, however the `session_state` does not actually change until the page is reloaded a _second_ time**. We discovered this using the `streamlit.sidebar` (however it behaves the same in any streamlit page.) We are using multiple pages in a dashboard (I.e. all our streamlit pages are in `streamlit_root/pages`) however we use a common sidebar to select what data is processed on the page. Previously, **if a user went to a new page, all selections on the sidebar were lost** - because the sidebar reloads from scratch. So we are attempting to use `session_state` to capture those selections, so they persist amongst all the pages. We did this by using the saved parameters within `session_state`, and then using the `session_state` to set the "default" parameter in the multiselect box (see code below.) However, this resulted in unpredictable behavior (see image.) I was able to work around, by using `experimental_rerun()` to force a second page reload anytime the multiselect was changed - but it causes a lot of slowdown and is fragile. **_Is it possible for the `session_state`` to be immediately updated after selecting from a multiselect box with a single page reload?_**  ### Reproducible Code Example ```Python import streamlit as st options = ["Apple", "Orange", "Pear"] # Initialize (check if session_state var exists and, if not, create it) try: garbage = st.session_state["multiselect_selections"] except KeyError: st.session_state["multiselect_selections"] = [] st.write('(BEFORE) st.session_state["multiselect_selections"]: ', st.session_state["multiselect_selections"]) st.session_state["multiselect_selections"] = st.multiselect("Multiselect", options, default = st.session_state["multiselect_selections"]) st.write('(AFTER) st.session_state["multiselect_selections"]: ', st.session_state["multiselect_selections"]) ``` ### Steps To Reproduce 1. streamlit run code.py 2. Select from the multiselect box. Now the "Before" and "After" state of the session_state variable 3. Select another options fro the multiselect box. 4. Not the selected option disappears, but the session_state variable is finally updated after this second page reload. ### Expected Behavior That session_state would be immediately (persistently) updated after the single page reload triggered by making a selection from the multiselect box. ### Current Behavior _No response_ ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: Streamlit, version 1.16.0 - Python version: 3.11 - Operating System: Windows 10 Enterprise - Browser: Chrome - Virtual environment: Na ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6079/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6078

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6078/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6078/comments

https://api.github.com/repos/streamlit/streamlit/issues/6078/events

https://github.com/streamlit/streamlit/issues/6078

I_kwDODCoeTs5d7OzD

File Uploader: Report failed uploads in multi-upload.

open

NONE

### Problem The problem is, that when doing a multi-upload using the file uploader widget, some files can fail sometimes (even though they are below the file size limit), see screenshot:  I tried uploading 13 files with ~12MB each. 7 failed and 6 succeeded. While errors don't happen regularly, it can happen. And the problem is, that the user cannot be informed about this, because I have no way to compare how many files the user intended to upload vs. how many actually uploaded. ### Solution **MVP:** What's the smallest possible solution that would get 80% of the problem out of the way? The obvious solution would be if the st.file_uploader object does not only create a list with the **successful** uploads, but also return an integer number how many files the user intended to upload in total. Then we can compare this with len of the list and decide whether any files failed or not. **Possible additions:** What are other things that could be added to the MVP over time to make it better? Even better would be of course to know, _which_ of the files failed (although not needed for me personally because I would just inform the user to redo the whole multi-upload). E.g. return a list object of the intended filenames that were supposed to upload. ### Additional context Code used: ``` uploaded_files = st.file_uploader("**.psdata** files", type=["psdata"], accept_multiple_files=True) for uploaded_file in uploaded_files: bytes_data = uploaded_file.read() st.write("filename:", uploaded_file.name, f"Bytes Len: {len(bytes_data)}") ``` --- Community voting on feature requests enables the Streamlit team to understand which features are most important to our users. **If you'd like the Streamlit team to prioritize this feature request, please use the 👍 (thumbs up emoji) reaction in response to the initial post.**

https://api.github.com/repos/streamlit/streamlit/issues/6078/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6077

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6077/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6077/comments

https://api.github.com/repos/streamlit/streamlit/issues/6077/events

https://github.com/streamlit/streamlit/pull/6077

PR_kwDODCoeTs5JfAPp

Fix Padding for Expander for fullScreenButton to show

closed

COLLABORATOR

<!-- Before contributing (PLEASE READ!) ⚠️ If your contribution is more than a few lines of code, then prior to starting to code on it please post in the issue saying you want to volunteer, then wait for a positive response. And if there is no issue for it yet, create it first. This helps make sure: 1. Two people aren't working on the same thing 2. This is something Streamlit's maintainers believe should be implemented/fixed 3. Any API, UI, or deeper architectural changes that need to be implemented have been fully thought through by Streamlit's maintainers 4. Your time is well spent! More information in our wiki: https://github.com/streamlit/streamlit/wiki/Contributing --> ## 📚 Context https://github.com/streamlit/streamlit/issues/5694 _Please describe the project or issue background here_ The issue is that the padding for the expander is too little and thus it doesn't show the full screen button. There are some solutions that I investigated such as passing in some props for `inExpander` and then modifying the fullScreenButton based on that prop but it didn't seem worth it to do all of the work to do that just so that we can see the button. I also updated the form because theyre similar elements / containers but form doesn't have the same problem. Note: this affects all the widgets / elements with full screen button that can go into the expander aka charts, dataframe, etc. - What kind of change does this PR introduce? - [x] Bugfix - [ ] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes Changing the padding - _Add bullet points summarizing your changes here_ - [ ] This is a breaking API change - [x] This is a visible (user-facing) change **Revised:** EXPANDER:  FORM:  _Insert screenshot of your updated UI/code here_ **Current:** EXPANDER:  FORM:  _Insert screenshot of existing UI/code here_ ## 🧪 Testing Done - [x] Screenshots included - [ ] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: Closes #5694 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6077/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6076

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6076/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6076/comments

https://api.github.com/repos/streamlit/streamlit/issues/6076/events

https://github.com/streamlit/streamlit/pull/6076

PR_kwDODCoeTs5Jd6Hv

Pre-commit check for print stmts

closed

COLLABORATOR

## 📚 Context Avoid rogue print statements in our code by adding pre-commit step to check for & error out if a print statement exists - What kind of change does this PR introduce? - [x] Other, please describe: Tech Debt/Update Pre-Commit Checks ## 🧠 Description of Changes - Added step to `.pre-commit-config.yaml` to check for print statements

https://api.github.com/repos/streamlit/streamlit/issues/6076/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6075

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6075/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6075/comments

https://api.github.com/repos/streamlit/streamlit/issues/6075/events

https://github.com/streamlit/streamlit/pull/6075

PR_kwDODCoeTs5Jd1VG

session_state: "common.py" module

closed

COLLABORATOR

Minor `streamlit.runtime.state` refactor: - There are a number of functions and types declared in `session_state.py` and shared between `session_state.py` and `widgets.py`. These are now moved to a new `common.py` module, so that we can avoid sibling modules importing from each other. - This also means that all functions that create, inspect, and validate widget_id values are now in the same source file. (This was the original motivation for the refactor.) - Renamed our `cull_nonexistent` functions, for removing stale widgets, to `remove_stale_widgets` - Renamed `get_widget_id` -> `compute_widget_id` - Renamed `GENERATED_WIDGET_KEY_PREFIX` -> `GENERATED_WIDGET_ID_PREFIX`, for consistency

https://api.github.com/repos/streamlit/streamlit/issues/6075/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6074

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6074/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6074/comments

https://api.github.com/repos/streamlit/streamlit/issues/6074/events

https://github.com/streamlit/streamlit/issues/6074

I_kwDODCoeTs5d2v0N

[Request for input] Keyed widget state persistence - discussion, possible fixes

open

NONE

👋 I work on developer experience for Streamlit and wanted to get some feedback on this recurring issue. **TL;DR: Take a look at the [example app](https://keyed-widget-state-issue.streamlit.app/) and let us know how it impacts you, and what you think about how to fix this issue!** **Discussion on the forums [here](https://discuss.streamlit.io/t/keyed-widget-state-persistence-discussion-possible-fixes/37359)** One of the current Streamlit behaviors that causes confusion for many of us is how widgets with `key='foo'` lose their session_state value when they are not rendered in a given run. It seems to come up most often with multipage apps that have some state dependency between pages, but also possible in a single page app with conditional rendering. A minimal example: ``` view = st.radio("View", ["view1", "view2"]) if view == "view1": st.text_input("Text1", key="text1") "☝️ Enter some text, then click on view2 above" elif view == "view2": "☝️ Now go back to view1 and see if your text is still there" ``` I'm working with @tconkling on whether / how to update this behavior so it's more intuitive, or at least possible to get the intuitive behavior easily. I put together a simple example app showing the behavior and linking some of the prior issues about it: [https://keyed-widget-state-issue.streamlit.app/](https://keyed-widget-state-issue.streamlit.app/) ### Questions - Do you have a full working public app where you had to hack around this (such as duplicating or using "shadow keys")? Please share with us! - Do you have an app or use case where you rely on the current behavior (widget has `key` set but needs value to reset after not rendering), that will break if we change this? We **really** want to hear about these! Please share! ### Possible fixes It seems like most developers we hear from find the current behavior very unintuitive, and we see a lot more use cases that have to work around this rather than benefit from it. - Given that, should we just do a breaking change to the default behavior to persist, like most of us seem to expect? - If so: We should still provide a path to clear keyed widget state when needed. Need to figure out how to best do this. If not a breaking change: most proposals seem to prefer something like a `persist=` key. For example, both @whitphx and @mathcatsand propose something like this in [#4458](https://github.com/streamlit/streamlit/issues/4458). Thoughts on this approach, or other ideas? --- Community voting on feature requests enables the Streamlit team to understand which features are most important to our users. **If you'd like the Streamlit team to prioritize this feature request, please use the 👍 (thumbs up emoji) reaction in response to the initial post.**

https://api.github.com/repos/streamlit/streamlit/issues/6074/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6073

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6073/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6073/comments

https://api.github.com/repos/streamlit/streamlit/issues/6073/events

https://github.com/streamlit/streamlit/issues/6073

I_kwDODCoeTs5d0sbU

Pandas SettingWithCopyWarning in type_util.py

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary type_util.py line 643 triggers a pandas SettingWithCopyWarning under certain condition. ### Reproducible Code Example [](https://issues.streamlitapp.com/?issue=gh-6073) ```Python import pandas as pd import streamlit as st df = pd.DataFrame({'a': [1, 2], 'b': [11, 22]}) print('okay with full df') st.dataframe(df) print('add a new column') df['value'] = [[1, 1.1], 2] print('problem with a small portion of df that includes the new column') st.dataframe(df[['a', 'value']]) print('no problem with the new full df') st.dataframe(df) if st.button('rerun'): print('---try experiment again---') st.experimental_rerun() ``` ### Steps To Reproduce Run the above code and see printouts for problem. Click button to rerun if desired. ### Expected Behavior There should not be any pandas SettingWithCopyWarning. ### Current Behavior SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame. Try using .loc[row_indexer,col_indexer] = value instead See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy df[col] = df[col].astype(str) ### Is this a regression? - [ ] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.16.0 - Python version: 3.9.6 - Operating System: Windows - Browser: Edge - Virtual environment: virtualenv ### Additional Information _No response_ ### Are you willing to submit a PR? - [ ] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6073/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6072

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6072/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6072/comments

https://api.github.com/repos/streamlit/streamlit/issues/6072/events

https://github.com/streamlit/streamlit/issues/6072

I_kwDODCoeTs5dx-nQ

Date range selector is hard to see in dark mode

open

NONE

### Checklist - [X] I have searched the [existing issues](https://github.com/streamlit/streamlit/issues) for similar issues. - [X] I added a very descriptive title to this issue. - [X] I have provided sufficient information below to help reproduce this issue. ### Summary When selecting multiple dates in dark mode, the range selector colors are very difficult to read. ### Reproducible Code Example ```Python import datetime import streamlit as st st.date_input("The date is", value=(datetime.datetime.now() - datetime.timedelta(days=30), datetime.datetime.now())) ``` ### Steps To Reproduce 1. Run the code above 2. Change the mode to `Dark Mode` while the app is running 3. Adjust the dates and see that the text in the dates is difficult to see. ### Expected Behavior 1. Run the code above 2. Change the mode to `Dark Mode` while the app is running 3. Adjust the dates and see that the text in the dates is black rather than white. ### Current Behavior The calendar dates are really hard to see.  ### Is this a regression? - [x] Yes, this used to work in a previous version. ### Debug info - Streamlit version: 1.17.0 - Python version: 3.10.9 - Operating System: MacOS 13.2 - Browser: Chrome - Virtual environment: Pipenv ### Additional Information _No response_ ### Are you willing to submit a PR? - [X] Yes, I am willing to submit a PR!

https://api.github.com/repos/streamlit/streamlit/issues/6072/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6071

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6071/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6071/comments

https://api.github.com/repos/streamlit/streamlit/issues/6071/events

https://github.com/streamlit/streamlit/pull/6071

PR_kwDODCoeTs5JYCxp

Add step parameter to st.time_input

open

CONTRIBUTOR

<!-- Before contributing (PLEASE READ!) ⚠️ If your contribution is more than a few lines of code, then prior to starting to code on it please post in the issue saying you want to volunteer, then wait for a positive response. And if there is no issue for it yet, create it first. This helps make sure: 1. Two people aren't working on the same thing 2. This is something Streamlit's maintainers believe should be implemented/fixed 3. Any API, UI, or deeper architectural changes that need to be implemented have been fully thought through by Streamlit's maintainers 4. Your time is well spent! More information in our wiki: https://github.com/streamlit/streamlit/wiki/Contributing --> ## 📚 Context _Please describe the project or issue background here_ - What kind of change does this PR introduce? This PR adds `step` parameter to `st.time_input` as a keyword only optional param. - [ ] Bugfix - [x] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - _Add bullet points summarizing your changes here_ - [ ] This is a breaking API change - [ ] This is a visible (user-facing) change **Revised:** _Insert screenshot of your updated UI/code here_ **Current:** _Insert screenshot of existing UI/code here_ ## 🧪 Testing Done - [ ] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: Closes #XXXX --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6071/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6070

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6070/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6070/comments

https://api.github.com/repos/streamlit/streamlit/issues/6070/events

https://github.com/streamlit/streamlit/pull/6070

PR_kwDODCoeTs5JXwVz

[DRAFT] Hamburger menu cleanup

closed

NONE

<!-- Before contributing (PLEASE READ!) ⚠️ If your contribution is more than a few lines of code, then prior to starting to code on it please post in the issue saying you want to volunteer, then wait for a positive response. And if there is no issue for it yet, create it first. This helps make sure: 1. Two people aren't working on the same thing 2. This is something Streamlit's maintainers believe should be implemented/fixed 3. Any API, UI, or deeper architectural changes that need to be implemented have been fully thought through by Streamlit's maintainers 4. Your time is well spent! More information in our wiki: https://github.com/streamlit/streamlit/wiki/Contributing --> ## 📚 Context _Please describe the project or issue background here_ This PR implements the hamburger menu items cleanup, according to the according to the "Clean up the hamburger menu" [Product spec](https://www.notion.so/streamlit/Product-spec-fa80e423ee5e4863963f5d0f4c873c7c). - What kind of change does this PR introduce? - [ ] Bugfix - [x] Feature - [x] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes * Renamed "About" to "About & help" menu item. * Moved "Report a bug" item from menu to the "About" modal * Moved "Get help" item from menu to the "About" modal * Removed the following menu items from "Developer options" section: * Streamlit could * Report a Streamlit bug * Visit Streamlit docs (link to documentation is now in the "About" modal) * Visit Streamlit forums (link to documentation is now in the "About" modal) * The "Promote this app" and "Report a bug with this app" menu items are added by the s4t project and removed in a [dedicated PR](https://github.com/streamlit/s4t/pull/5087) * Visual changes to the "About modal": * Information about Streamlit version and Streamlit-related urls are separated into a dedicated component * URLs are blue not red * Horizontal line separates user-added markdown from the core streamlit into Changes: - [ ] This is a breaking API change - [x] This is a visible (user-facing) change **Revised:** _Insert screenshot of your updated UI/code here_ *Simplified hamburger menu in the local environment: <img width="206" alt="Screenshot 2023-02-07 at 10 24 50" src="https://user-images.githubusercontent.com/119880369/217205261-d04ea783-f4bb-459e-8489-fae184a7487b.png"> * "About" modal with custom markdown message specified by the user: <img width="513" alt="Screenshot 2023-02-07 at 14 18 46" src="https://user-images.githubusercontent.com/119880369/217255777-188a2224-ca43-4912-a9c7-af21a130c162.png"> * "About" modal default view: <img width="520" alt="Screenshot 2023-02-07 at 10 26 02" src="https://user-images.githubusercontent.com/119880369/217205583-2434827a-4236-4f48-8814-6a814e72eb41.png"> **Current:** _Insert screenshot of existing UI/code here_ * Original hamburger menu in the local environment: <img width="222" alt="Screenshot 2023-02-07 at 10 31 20" src="https://user-images.githubusercontent.com/119880369/217206679-f12498cd-7d48-47c0-9c07-19f260e53ee2.png"> * Original "About" modal with custom markdown message specified by the user: <img width="513" alt="Screenshot 2023-02-07 at 10 32 10" src="https://user-images.githubusercontent.com/119880369/217206759-c5796dd9-c3e1-4092-9ced-5678b0d9ee73.png"> * Original "About" modal default view: <img width="510" alt="Screenshot 2023-02-07 at 10 31 35" src="https://user-images.githubusercontent.com/119880369/217206799-6e155917-b1a8-4c39-8757-e449199657e6.png"> ## 🧪 Testing Done - [x] Screenshots included - [x] Added/Updated unit tests - [ ] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: [STREAMLIT-2113](https://snowflakecomputing.atlassian.net/browse/STREAMLIT-2113) - **PR in s4t repo**: https://github.com/streamlit/s4t/pull/5087 --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6070/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6069

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6069/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6069/comments

https://api.github.com/repos/streamlit/streamlit/issues/6069/events

https://github.com/streamlit/streamlit/pull/6069

PR_kwDODCoeTs5JXQSd

Add sfc-gh-mnowotka as a protobuf CODEOWNER

closed

COLLABORATOR

Protobuf changes are common, and we'd like the Warsaw team to be able to approve them

https://api.github.com/repos/streamlit/streamlit/issues/6069/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6068

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6068/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6068/comments

https://api.github.com/repos/streamlit/streamlit/issues/6068/events

https://github.com/streamlit/streamlit/issues/6068

I_kwDODCoeTs5dwLmb

Image Carousel

open

NONE

I know there have been "official workarounds" in the past but it seems like none of them exist or work anymore. This feature has already been requested: Here: https://discuss.streamlit.io/t/feature-suggestion-image-carousel/1189 and here: https://github.com/streamlit/streamlit/issues/780 it seems like people have used image carousels written in react to make this work but there's no official streamlit solution.

https://api.github.com/repos/streamlit/streamlit/issues/6068/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6067

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6067/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6067/comments

https://api.github.com/repos/streamlit/streamlit/issues/6067/events

https://github.com/streamlit/streamlit/pull/6067

PR_kwDODCoeTs5JVZPd

Allow control of st.pyplot() width

open

CONTRIBUTOR

<!-- Before contributing (PLEASE READ!) ⚠️ If your contribution is more than a few lines of code, then prior to starting to code on it please post in the issue saying you want to volunteer, then wait for a positive response. And if there is no issue for it yet, create it first. This helps make sure: 1. Two people aren't working on the same thing 2. This is something Streamlit's maintainers believe should be implemented/fixed 3. Any API, UI, or deeper architectural changes that need to be implemented have been fully thought through by Streamlit's maintainers 4. Your time is well spent! More information in our wiki: https://github.com/streamlit/streamlit/wiki/Contributing --> ## 📚 Context Close: https://github.com/streamlit/streamlit/issues/3527 _Please describe the project or issue background here_ - What kind of change does this PR introduce? - [ ] Bugfix - [X] Feature - [ ] Refactoring - [ ] Other, please describe: ## 🧠 Description of Changes - _Add bullet points summarizing your changes here_ - [ ] This is a breaking API change - [ ] This is a visible (user-facing) change **Revised:** _Insert screenshot of your updated UI/code here_ **Current:** _Insert screenshot of existing UI/code here_ ## 🧪 Testing Done - [ ] Screenshots included - [X] Added/Updated unit tests - [X] Added/Updated e2e tests ## 🌐 References _Does this depend on other work, documents, or tickets?_ - **Issue**: Closes #XXXX --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

https://api.github.com/repos/streamlit/streamlit/issues/6067/timeline

https://api.github.com/repos/streamlit/streamlit/issues/6066

https://api.github.com/repos/streamlit/streamlit

https://api.github.com/repos/streamlit/streamlit/issues/6066/labels{/name}

https://api.github.com/repos/streamlit/streamlit/issues/6066/comments

https://api.github.com/repos/streamlit/streamlit/issues/6066/events

https://github.com/streamlit/streamlit/issues/6066

I_kwDODCoeTs5dspIN

Streamlit startup time could be reduced from 1s to 400ms