datasets
Browse files- create_dataset_from_tsv.py +53 -0
- tmx2cvs.py → tmx2tsv.py +0 -0
- tsv_all_source_target/dev.tsv +3 -0
- tsv_all_source_target/test.tsv +3 -0
- tsv_all_source_target/train.tsv +3 -0
create_dataset_from_tsv.py
ADDED
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import pandas as pd
|
2 |
+
import xmltodict
|
3 |
+
from sklearn.model_selection import train_test_split
|
4 |
+
import glob
|
5 |
+
import sys
|
6 |
+
import os
|
7 |
+
|
8 |
+
filelist = glob.glob('tsv_source_target/*.tsv')
|
9 |
+
|
10 |
+
data = pd.DataFrame()
|
11 |
+
|
12 |
+
for tsvfile in filelist:
|
13 |
+
tmp = pd.read_csv(tsvfile, sep='\t')
|
14 |
+
tmp.columns=['source','target']
|
15 |
+
tmp['rev_source'] = tmp['target']
|
16 |
+
tmp['rev_target'] = tmp['source']
|
17 |
+
|
18 |
+
|
19 |
+
path = tsvfile.split("/")
|
20 |
+
source = path[1][0:3]
|
21 |
+
target = path[1][3:6]
|
22 |
+
|
23 |
+
prefix = f"{source}_{target}: "
|
24 |
+
tmp['source'] = prefix + tmp['source']
|
25 |
+
|
26 |
+
rev_prefix = f"{target}_{source}: "
|
27 |
+
tmp['rev_source'] = rev_prefix + tmp['rev_source']
|
28 |
+
|
29 |
+
data = pd.concat([data,tmp])
|
30 |
+
|
31 |
+
|
32 |
+
#Shuffle
|
33 |
+
data = data.sample(frac=1).reset_index(drop=True)
|
34 |
+
|
35 |
+
# Add both directions
|
36 |
+
original = data[['source','target']]
|
37 |
+
reverse = data[['rev_source','rev_target']]
|
38 |
+
reverse.columns=['source','target']
|
39 |
+
|
40 |
+
data = pd.concat([original,reverse])
|
41 |
+
data = data.sample(frac=1).reset_index(drop=True)
|
42 |
+
|
43 |
+
# Train - test - dev
|
44 |
+
train, test = train_test_split(data, test_size=0.2)
|
45 |
+
test, dev = train_test_split(test, test_size=0.5)
|
46 |
+
|
47 |
+
# Write the datasets to disk
|
48 |
+
train.to_csv('tsv_all_source_target/train.tsv', index=False, header=False, sep='\t')
|
49 |
+
test.to_csv('tsv_all_source_target/test.tsv', index=False, header=False, sep='\t')
|
50 |
+
dev.to_csv('tsv_all_source_target/dev.tsv', index=False, header=False, sep='\t')
|
51 |
+
|
52 |
+
|
53 |
+
print("Finished")
|
tmx2cvs.py → tmx2tsv.py
RENAMED
File without changes
|
tsv_all_source_target/dev.tsv
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f5de7e7fc05d9697ff9863f4846364878492e847d236c33d4e55495c66b3c2ae
|
3 |
+
size 25212526
|
tsv_all_source_target/test.tsv
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:53096012b9e502826ee0028367316aa624d6a1b847eb255f173b09d9475af4ce
|
3 |
+
size 25054929
|
tsv_all_source_target/train.tsv
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1f28b52818b79814f12c32e2ddf11f434e52439497a11bb8f251a89d682ac91e
|
3 |
+
size 201104421
|