|
import pandas as pd |
|
import xmltodict |
|
from sklearn.model_selection import train_test_split |
|
import glob |
|
import sys |
|
import os |
|
|
|
filelist = glob.glob('translate_dataset/*.tsv') |
|
|
|
data = pd.DataFrame() |
|
|
|
for tsvfile in filelist: |
|
tmp = pd.read_csv(tsvfile, sep='\t') |
|
tmp.columns=['source','target'] |
|
tmp['rev_source'] = tmp['target'] |
|
tmp['rev_target'] = tmp['source'] |
|
|
|
|
|
path = tsvfile.split("/") |
|
source = path[1][0:3] |
|
target = path[1][3:6] |
|
|
|
prefix = f"{source}_{target}: " |
|
tmp['source'] = prefix + tmp['source'] |
|
|
|
rev_prefix = f"{target}_{source}: " |
|
tmp['rev_source'] = rev_prefix + tmp['rev_source'] |
|
|
|
data = pd.concat([data,tmp]) |
|
|
|
|
|
|
|
data = data.sample(frac=1).reset_index(drop=True) |
|
|
|
|
|
original = data[['source','target']] |
|
reverse = data[['rev_source','rev_target']] |
|
reverse.columns=['source','target'] |
|
|
|
data = pd.concat([original,reverse]) |
|
|
|
data['source'] = data['source'].str[4:] |
|
|
|
data = data.sample(frac=1).reset_index(drop=True) |
|
|
|
|
|
data['source'] = data['source'].str.replace('\t',' ') |
|
data['source'] = data['source'].str.replace('\n',' ') |
|
data['target'] = data['target'].str.replace('\t',' ') |
|
data['target'] = data['target'].str.replace('\n',' ') |
|
|
|
|
|
|
|
|
|
train, test = train_test_split(data, test_size=0.2) |
|
test, dev = train_test_split(test, test_size=0.5) |
|
|
|
|
|
test.to_csv('tsv_all_target/test.tsv', index=False, header=False, sep='\t') |
|
dev.to_csv('tsv_all_target/dev.tsv', index=False, header=False, sep='\t') |
|
print("Finished writint dev and test. Now we are adding some extra language detection and English translation") |
|
|
|
|
|
filelist = glob.glob('langid/langid_datafiles/*train.tsv') |
|
|
|
for tsvfile in filelist: |
|
tmp = pd.read_csv(tsvfile, sep='\t') |
|
tmp.columns=['source','target'] |
|
|
|
train = pd.concat([train,tmp]) |
|
|
|
|
|
|
|
tmp = pd.read_csv('eng_nob/train.tsv', sep='\t') |
|
tmp.columns=['source','target'] |
|
train = pd.concat([train,tmp]) |
|
|
|
|
|
train = train.sample(frac=1).reset_index(drop=True) |
|
|
|
|
|
train['source'] = train['source'].str.replace('\t',' ') |
|
train['source'] = train['source'].str.replace('\n',' ') |
|
train['target'] = train['target'].str.replace('\t',' ') |
|
train['target'] = train['target'].str.replace('\n',' ') |
|
|
|
|
|
train.to_csv('tsv_all_target/train.tsv', index=False, header=False, sep='\t') |
|
|
|
print("Finished") |
|
|