-
Notifications
You must be signed in to change notification settings - Fork 268
/
data.html
1063 lines (990 loc) · 59.6 KB
/
data.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1, minimum-scale=1" />
<meta name="generator" content="pdoc 0.9.2" />
<title>ktrain.text.data API documentation</title>
<meta name="description" content="" />
<link rel="preload stylesheet" as="style" href="https://cdnjs.cloudflare.com/ajax/libs/10up-sanitize.css/11.0.1/sanitize.min.css" integrity="sha256-PK9q560IAAa6WVRRh76LtCaI8pjTJ2z11v0miyNNjrs=" crossorigin>
<link rel="preload stylesheet" as="style" href="https://cdnjs.cloudflare.com/ajax/libs/10up-sanitize.css/11.0.1/typography.min.css" integrity="sha256-7l/o7C8jubJiy74VsKTidCy1yBkRtiUGbVkYBylBqUg=" crossorigin>
<link rel="stylesheet preload" as="style" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/10.1.1/styles/github.min.css" crossorigin>
<style>:root{--highlight-color:#fe9}.flex{display:flex !important}body{line-height:1.5em}#content{padding:20px}#sidebar{padding:30px;overflow:hidden}#sidebar > *:last-child{margin-bottom:2cm}.http-server-breadcrumbs{font-size:130%;margin:0 0 15px 0}#footer{font-size:.75em;padding:5px 30px;border-top:1px solid #ddd;text-align:right}#footer p{margin:0 0 0 1em;display:inline-block}#footer p:last-child{margin-right:30px}h1,h2,h3,h4,h5{font-weight:300}h1{font-size:2.5em;line-height:1.1em}h2{font-size:1.75em;margin:1em 0 .50em 0}h3{font-size:1.4em;margin:25px 0 10px 0}h4{margin:0;font-size:105%}h1:target,h2:target,h3:target,h4:target,h5:target,h6:target{background:var(--highlight-color);padding:.2em 0}a{color:#058;text-decoration:none;transition:color .3s ease-in-out}a:hover{color:#e82}.title code{font-weight:bold}h2[id^="header-"]{margin-top:2em}.ident{color:#900}pre code{background:#f8f8f8;font-size:.8em;line-height:1.4em}code{background:#f2f2f1;padding:1px 4px;overflow-wrap:break-word}h1 code{background:transparent}pre{background:#f8f8f8;border:0;border-top:1px solid #ccc;border-bottom:1px solid #ccc;margin:1em 0;padding:1ex}#http-server-module-list{display:flex;flex-flow:column}#http-server-module-list div{display:flex}#http-server-module-list dt{min-width:10%}#http-server-module-list p{margin-top:0}.toc ul,#index{list-style-type:none;margin:0;padding:0}#index code{background:transparent}#index h3{border-bottom:1px solid #ddd}#index ul{padding:0}#index h4{margin-top:.6em;font-weight:bold}@media (min-width:200ex){#index .two-column{column-count:2}}@media (min-width:300ex){#index .two-column{column-count:3}}dl{margin-bottom:2em}dl dl:last-child{margin-bottom:4em}dd{margin:0 0 1em 3em}#header-classes + dl > dd{margin-bottom:3em}dd dd{margin-left:2em}dd p{margin:10px 0}.name{background:#eee;font-weight:bold;font-size:.85em;padding:5px 10px;display:inline-block;min-width:40%}.name:hover{background:#e0e0e0}dt:target .name{background:var(--highlight-color)}.name > span:first-child{white-space:nowrap}.name.class > span:nth-child(2){margin-left:.4em}.inherited{color:#999;border-left:5px solid #eee;padding-left:1em}.inheritance em{font-style:normal;font-weight:bold}.desc h2{font-weight:400;font-size:1.25em}.desc h3{font-size:1em}.desc dt code{background:inherit}.source summary,.git-link-div{color:#666;text-align:right;font-weight:400;font-size:.8em;text-transform:uppercase}.source summary > *{white-space:nowrap;cursor:pointer}.git-link{color:inherit;margin-left:1em}.source pre{max-height:500px;overflow:auto;margin:0}.source pre code{font-size:12px;overflow:visible}.hlist{list-style:none}.hlist li{display:inline}.hlist li:after{content:',\2002'}.hlist li:last-child:after{content:none}.hlist .hlist{display:inline;padding-left:1em}img{max-width:100%}td{padding:0 .5em}.admonition{padding:.1em .5em;margin-bottom:1em}.admonition-title{font-weight:bold}.admonition.note,.admonition.info,.admonition.important{background:#aef}.admonition.todo,.admonition.versionadded,.admonition.tip,.admonition.hint{background:#dfd}.admonition.warning,.admonition.versionchanged,.admonition.deprecated{background:#fd4}.admonition.error,.admonition.danger,.admonition.caution{background:lightpink}</style>
<style media="screen and (min-width: 700px)">@media screen and (min-width:700px){#sidebar{width:30%;height:100vh;overflow:auto;position:sticky;top:0}#content{width:70%;max-width:100ch;padding:3em 4em;border-left:1px solid #ddd}pre code{font-size:1em}.item .name{font-size:1em}main{display:flex;flex-direction:row-reverse;justify-content:flex-end}.toc ul ul,#index ul{padding-left:1.5em}.toc > ul > li{margin-top:.5em}}</style>
<style media="print">@media print{#sidebar h1{page-break-before:always}.source{display:none}}@media print{*{background:transparent !important;color:#000 !important;box-shadow:none !important;text-shadow:none !important}a[href]:after{content:" (" attr(href) ")";font-size:90%}a[href][title]:after{content:none}abbr[title]:after{content:" (" attr(title) ")"}.ir a:after,a[href^="javascript:"]:after,a[href^="#"]:after{content:""}pre,blockquote{border:1px solid #999;page-break-inside:avoid}thead{display:table-header-group}tr,img{page-break-inside:avoid}img{max-width:100% !important}@page{margin:0.5cm}p,h2,h3{orphans:3;widows:3}h1,h2,h3,h4,h5,h6{page-break-after:avoid}}</style>
<script defer src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/10.1.1/highlight.min.js" integrity="sha256-Uv3H6lx7dJmRfRvH8TH6kJD1TSK1aFcwgx+mdg3epi8=" crossorigin></script>
<script>window.addEventListener('DOMContentLoaded', () => hljs.initHighlighting())</script>
</head>
<body>
<main>
<article id="content">
<header>
<h1 class="title">Module <code>ktrain.text.data</code></h1>
</header>
<section id="section-intro">
<details class="source">
<summary>
<span>Expand source code</span>
</summary>
<pre><code class="python">from ..imports import *
from .. import utils as U
from . import preprocessor as tpp
from . import textutils as TU
MAX_FEATURES = 20000
MAXLEN = 400
def texts_from_folder(datadir, classes=None,
max_features=MAX_FEATURES, maxlen=MAXLEN,
ngram_range=1,
train_test_names=['train', 'test'],
preprocess_mode='standard',
encoding=None, # detected automatically
lang=None, # detected automatically
val_pct=0.1, random_state=None,
verbose=1):
"""
```
Returns corpus as sequence of word IDs.
Assumes corpus is in the following folder structure:
├── datadir
│ ├── train
│ │ ├── class0 # folder containing documents of class 0
│ │ ├── class1 # folder containing documents of class 1
│ │ ├── class2 # folder containing documents of class 2
│ │ └── classN # folder containing documents of class N
│ └── test
│ ├── class0 # folder containing documents of class 0
│ ├── class1 # folder containing documents of class 1
│ ├── class2 # folder containing documents of class 2
│ └── classN # folder containing documents of class N
Each subfolder should contain documents in plain text format.
If train and test contain additional subfolders that do not represent
classes, they can be ignored by explicitly listing the subfolders of
interest using the classes argument.
Args:
datadir (str): path to folder
classes (list): list of classes (subfolders to consider).
This is simply supplied as the categories argument
to sklearn's load_files function.
max_features (int): maximum number of unigrams to consider
Note: This is only used for preprocess_mode='standard'.
maxlen (int): maximum length of tokens in document
ngram_range (int): If > 1, will include 2=bigrams, 3=trigrams and bigrams
train_test_names (list): list of strings represnting the subfolder
name for train and validation sets
if test name is missing, <val_pct> of training
will be used for validation
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
encoding (str): character encoding to use. Auto-detected if None
lang (str): language. Auto-detected if None.
val_pct(float): Onlyl used if train_test_names has 1 and not 2 names
random_state(int): If integer is supplied, train/test split is reproducible.
IF None, train/test split will be random
verbose (bool): verbosity
```
"""
# check train_test_names
if len(train_test_names) < 1 or len(train_test_names) > 2:
raise ValueError('train_test_names must have 1 or two elements for train and optionally validation')
# read in training and test corpora
train_str = train_test_names[0]
train_b = load_files(os.path.join(datadir, train_str), shuffle=True, categories=classes)
if len(train_test_names) > 1:
test_str = train_test_names[1]
test_b = load_files(os.path.join(datadir, test_str), shuffle=False, categories=classes)
x_train = train_b.data
y_train = train_b.target
x_test = test_b.data
y_test = test_b.target
else:
x_train, x_test, y_train, y_test = train_test_split(train_b.data,
train_b.target,
test_size=val_pct,
random_state=random_state)
# decode based on supplied encoding
if encoding is None:
encoding = TU.detect_encoding(x_train)
U.vprint('detected encoding: %s' % (encoding), verbose=verbose)
try:
x_train = [x.decode(encoding) for x in x_train]
x_test = [x.decode(encoding) for x in x_test]
except:
U.vprint('Decoding with %s failed 1st attempt - using %s with skips' % (encoding,
encoding),
verbose=verbose)
x_train = TU.decode_by_line(x_train, encoding=encoding, verbose=verbose)
x_test = TU.decode_by_line(x_test, encoding=encoding, verbose=verbose)
# detect language
if lang is None: lang = TU.detect_lang(x_train)
check_unsupported_lang(lang, preprocess_mode)
# return preprocessed the texts
preproc_type = tpp.TEXT_PREPROCESSORS.get(preprocess_mode, None)
if None: raise ValueError('unsupported preprocess_mode')
preproc = preproc_type(maxlen,
max_features,
class_names = train_b.target_names,
lang=lang, ngram_range=ngram_range)
trn = preproc.preprocess_train(x_train, y_train, verbose=verbose)
val = preproc.preprocess_test(x_test, y_test, verbose=verbose)
return (trn, val, preproc)
def texts_from_csv(train_filepath,
text_column,
label_columns = [],
val_filepath=None,
max_features=MAX_FEATURES, maxlen=MAXLEN,
val_pct=0.1, ngram_range=1, preprocess_mode='standard',
encoding=None, # auto-detected
lang=None, # auto-detected
sep=',',
is_regression=False,
random_state=None,
verbose=1):
"""
```
Loads text data from CSV or TSV file. Class labels are assumed to be
one of the following formats:
1. one-hot-encoded or multi-hot-encoded arrays representing classes:
Example with label_columns=['positive', 'negative'] and text_column='text':
text|positive|negative
I like this movie.|1|0
I hated this movie.|0|1
Classification will have a single one in each row: [[1,0,0], [0,1,0]]]
Multi-label classification will have one more ones in each row: [[1,1,0], [0,1,1]]
2. labels are in a single column of string or integer values representing classs labels
Example with label_columns=['label'] and text_column='text':
text|label
I like this movie.|positive
I hated this movie.|negative
3. labels are a single column of numerical values for text regression
NOTE: Must supply is_regression=True for labels to be treated as numerical targets
wine_description|wine_price
Exquisite wine!|100
Wine for budget shoppers|8
Args:
train_filepath(str): file path to training CSV
text_column(str): name of column containing the text
label_column(list): list of columns that are to be treated as labels
val_filepath(string): file path to test CSV. If not supplied,
10% of documents in training CSV will be
used for testing/validation.
max_features(int): max num of words to consider in vocabulary
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if val_filepath is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
encoding (str): character encoding to use. Auto-detected if None
lang (str): language. Auto-detected if None.
sep(str): delimiter for CSV (comma is default)
is_regression(bool): If True, integer targets will be treated as numerical targets instead of class IDs
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random
verbose (boolean): verbosity
```
"""
if encoding is None:
with open(train_filepath, 'rb') as f:
#encoding = chardet.detect(f.read())['encoding']
#encoding = 'utf-8' if encoding.lower() in ['ascii', 'utf8', 'utf-8'] else encoding
encoding = TU.detect_encoding(f.read())
U.vprint('detected encoding: %s (if wrong, set manually)' % (encoding), verbose=verbose)
train_df = pd.read_csv(train_filepath, encoding=encoding,sep=sep)
val_df = pd.read_csv(val_filepath, encoding=encoding,sep=sep) if val_filepath is not None else None
return texts_from_df(train_df,
text_column,
label_columns=label_columns,
val_df = val_df,
max_features=max_features,
maxlen=maxlen,
val_pct=val_pct,
ngram_range=ngram_range,
preprocess_mode=preprocess_mode,
lang=lang, is_regression=is_regression, random_state=random_state,
verbose=verbose)
def texts_from_df(train_df,
text_column,
label_columns = [],
val_df=None,
max_features=MAX_FEATURES, maxlen=MAXLEN,
val_pct=0.1, ngram_range=1, preprocess_mode='standard',
lang=None, # auto-detected
is_regression=False,
random_state=None,
verbose=1):
"""
```
Loads text data from Pandas dataframe file. Class labels are assumed to be
one of the following formats:
1. one-hot-encoded or multi-hot-encoded arrays representing classes:
Example with label_columns=['positive', 'negative'] and text_column='text':
text|positive|negative
I like this movie.|1|0
I hated this movie.|0|1
Classification will have a single one in each row: [[1,0,0], [0,1,0]]]
Multi-label classification will have one more ones in each row: [[1,1,0], [0,1,1]]
2. labels are in a single column of string or integer values representing class labels
Example with label_columns=['label'] and text_column='text':
text|label
I like this movie.|positive
I hated this movie.|negative
3. labels are a single column of numerical values for text regression
NOTE: Must supply is_regression=True for integer labels to be treated as numerical targets
wine_description|wine_price
Exquisite wine!|100
Wine for budget shoppers|8
Args:
train_df(dataframe): Pandas dataframe
text_column(str): name of column containing the text
label_columns(list): list of columns that are to be treated as labels
val_df(dataframe): file path to test dataframe. If not supplied,
10% of documents in training df will be
used for testing/validation.
max_features(int): max num of words to consider in vocabulary.
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if val_filepath is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
lang (str): language. Auto-detected if None.
is_regression(bool): If True, integer targets will be treated as numerical targets instead of class IDs
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random
verbose (boolean): verbosity
```
"""
# read in train and test data
train_df = train_df.copy()
train_df[text_column].fillna('fillna', inplace=True)
if val_df is not None:
val_df = val_df.copy()
val_df[text_column].fillna('fillna', inplace=True)
else:
train_df, val_df = train_test_split(train_df, test_size=val_pct, random_state=random_state)
# transform labels
ytransdf = U.YTransformDataFrame(label_columns, is_regression=is_regression)
t_df = ytransdf.apply_train(train_df)
v_df = ytransdf.apply_test(val_df)
class_names = ytransdf.get_classes()
new_lab_cols = ytransdf.get_label_columns(squeeze=True)
x_train = t_df[text_column].values
y_train = t_df[new_lab_cols].values
x_test = v_df[text_column].values
y_test = v_df[new_lab_cols].values
# detect language
if lang is None: lang = TU.detect_lang(x_train)
check_unsupported_lang(lang, preprocess_mode)
# return preprocessed the texts
preproc_type = tpp.TEXT_PREPROCESSORS.get(preprocess_mode, None)
if None: raise ValueError('unsupported preprocess_mode')
preproc = preproc_type(maxlen,
max_features,
class_names = class_names,
lang=lang, ngram_range=ngram_range)
trn = preproc.preprocess_train(x_train, y_train, verbose=verbose)
val = preproc.preprocess_test(x_test, y_test, verbose=verbose)
# QUICKFIX for #314
preproc.ytransform.le = ytransdf.le
return (trn, val, preproc)
def texts_from_array(x_train, y_train, x_test=None, y_test=None,
class_names = [],
max_features=MAX_FEATURES, maxlen=MAXLEN,
val_pct=0.1, ngram_range=1, preprocess_mode='standard',
lang=None, # auto-detected
random_state=None,
verbose=1):
"""
```
Loads and preprocesses text data from arrays.
texts_from_array can handle data for both text classification
and text regression. If class_names is empty, a regression task is assumed.
Args:
x_train(list): list of training texts
y_train(list): labels in one of the following forms:
1. list of integers representing classes (class_names is required)
2. list of strings representing classes (class_names is not needed and ignored.)
3. a one or multi hot encoded array representing classes (class_names is required)
4. numerical values for text regresssion (class_names should be left empty)
x_test(list): list of training texts
y_test(list): labels in one of the following forms:
1. list of integers representing classes (class_names is required)
2. list of strings representing classes (class_names is not needed and ignored.)
3. a one or multi hot encoded array representing classes (class_names is required)
4. numerical values for text regresssion (class_names should be left empty)
class_names (list): list of strings representing class labels
shape should be (num_examples,1) or (num_examples,)
max_features(int): max num of words to consider in vocabulary
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if x_val and y_val is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
lang (str): language. Auto-detected if None.
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random.
verbose (boolean): verbosity
```
"""
U.check_array(x_train, y=y_train, X_name='x_train', y_name='y_train')
if x_test is None or y_test is None:
x_train, x_test, y_train, y_test = train_test_split(x_train, y_train,
test_size=val_pct,
random_state=random_state)
else:
U.check_array(x_test, y=y_test, X_name='x_test', y_name='y_test')
# removed as TextPreprocessor now handles this.
#if isinstance(y_train[0], str):
#if not isinstance(y_test[0], str):
#raise ValueError('y_train contains strings, but y_test does not')
#encoder = LabelEncoder()
#encoder.fit(y_train)
#y_train = encoder.transform(y_train)
#y_test = encoder.transform(y_test)
# detect language
if lang is None: lang = TU.detect_lang(x_train)
check_unsupported_lang(lang, preprocess_mode)
# return preprocessed the texts
preproc_type = tpp.TEXT_PREPROCESSORS.get(preprocess_mode, None)
if None: raise ValueError('unsupported preprocess_mode')
preproc = preproc_type(maxlen,
max_features,
class_names = class_names,
lang=lang, ngram_range=ngram_range)
trn = preproc.preprocess_train(x_train, y_train, verbose=verbose)
val = preproc.preprocess_test(x_test, y_test, verbose=verbose)
if not preproc.get_classes() and verbose:
print('task: text regression (supply class_names argument if this is supposed to be classification task)')
else:
print('task: text classification')
return (trn, val, preproc)
def check_unsupported_lang(lang, preprocess_mode):
"""
```
check for unsupported language (e.g., nospace langs not supported by Jieba)
```
"""
unsupported = preprocess_mode=='standard' and TU.is_nospace_lang(lang) and not TU.is_chinese(lang)
if unsupported:
raise ValueError('language %s is currently only supported by the BERT model. ' % (lang) +
'Please select preprocess_mode="bert"')</code></pre>
</details>
</section>
<section>
</section>
<section>
</section>
<section>
<h2 class="section-title" id="header-functions">Functions</h2>
<dl>
<dt id="ktrain.text.data.check_unsupported_lang"><code class="name flex">
<span>def <span class="ident">check_unsupported_lang</span></span>(<span>lang, preprocess_mode)</span>
</code></dt>
<dd>
<div class="desc"><pre><code>check for unsupported language (e.g., nospace langs not supported by Jieba)
</code></pre></div>
<details class="source">
<summary>
<span>Expand source code</span>
</summary>
<pre><code class="python">def check_unsupported_lang(lang, preprocess_mode):
"""
```
check for unsupported language (e.g., nospace langs not supported by Jieba)
```
"""
unsupported = preprocess_mode=='standard' and TU.is_nospace_lang(lang) and not TU.is_chinese(lang)
if unsupported:
raise ValueError('language %s is currently only supported by the BERT model. ' % (lang) +
'Please select preprocess_mode="bert"')</code></pre>
</details>
</dd>
<dt id="ktrain.text.data.texts_from_array"><code class="name flex">
<span>def <span class="ident">texts_from_array</span></span>(<span>x_train, y_train, x_test=None, y_test=None, class_names=[], max_features=20000, maxlen=400, val_pct=0.1, ngram_range=1, preprocess_mode='standard', lang=None, random_state=None, verbose=1)</span>
</code></dt>
<dd>
<div class="desc"><pre><code>Loads and preprocesses text data from arrays.
texts_from_array can handle data for both text classification
and text regression. If class_names is empty, a regression task is assumed.
Args:
x_train(list): list of training texts
y_train(list): labels in one of the following forms:
1. list of integers representing classes (class_names is required)
2. list of strings representing classes (class_names is not needed and ignored.)
3. a one or multi hot encoded array representing classes (class_names is required)
4. numerical values for text regresssion (class_names should be left empty)
x_test(list): list of training texts
y_test(list): labels in one of the following forms:
1. list of integers representing classes (class_names is required)
2. list of strings representing classes (class_names is not needed and ignored.)
3. a one or multi hot encoded array representing classes (class_names is required)
4. numerical values for text regresssion (class_names should be left empty)
class_names (list): list of strings representing class labels
shape should be (num_examples,1) or (num_examples,)
max_features(int): max num of words to consider in vocabulary
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if x_val and y_val is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
lang (str): language. Auto-detected if None.
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random.
verbose (boolean): verbosity
</code></pre></div>
<details class="source">
<summary>
<span>Expand source code</span>
</summary>
<pre><code class="python">def texts_from_array(x_train, y_train, x_test=None, y_test=None,
class_names = [],
max_features=MAX_FEATURES, maxlen=MAXLEN,
val_pct=0.1, ngram_range=1, preprocess_mode='standard',
lang=None, # auto-detected
random_state=None,
verbose=1):
"""
```
Loads and preprocesses text data from arrays.
texts_from_array can handle data for both text classification
and text regression. If class_names is empty, a regression task is assumed.
Args:
x_train(list): list of training texts
y_train(list): labels in one of the following forms:
1. list of integers representing classes (class_names is required)
2. list of strings representing classes (class_names is not needed and ignored.)
3. a one or multi hot encoded array representing classes (class_names is required)
4. numerical values for text regresssion (class_names should be left empty)
x_test(list): list of training texts
y_test(list): labels in one of the following forms:
1. list of integers representing classes (class_names is required)
2. list of strings representing classes (class_names is not needed and ignored.)
3. a one or multi hot encoded array representing classes (class_names is required)
4. numerical values for text regresssion (class_names should be left empty)
class_names (list): list of strings representing class labels
shape should be (num_examples,1) or (num_examples,)
max_features(int): max num of words to consider in vocabulary
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if x_val and y_val is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
lang (str): language. Auto-detected if None.
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random.
verbose (boolean): verbosity
```
"""
U.check_array(x_train, y=y_train, X_name='x_train', y_name='y_train')
if x_test is None or y_test is None:
x_train, x_test, y_train, y_test = train_test_split(x_train, y_train,
test_size=val_pct,
random_state=random_state)
else:
U.check_array(x_test, y=y_test, X_name='x_test', y_name='y_test')
# removed as TextPreprocessor now handles this.
#if isinstance(y_train[0], str):
#if not isinstance(y_test[0], str):
#raise ValueError('y_train contains strings, but y_test does not')
#encoder = LabelEncoder()
#encoder.fit(y_train)
#y_train = encoder.transform(y_train)
#y_test = encoder.transform(y_test)
# detect language
if lang is None: lang = TU.detect_lang(x_train)
check_unsupported_lang(lang, preprocess_mode)
# return preprocessed the texts
preproc_type = tpp.TEXT_PREPROCESSORS.get(preprocess_mode, None)
if None: raise ValueError('unsupported preprocess_mode')
preproc = preproc_type(maxlen,
max_features,
class_names = class_names,
lang=lang, ngram_range=ngram_range)
trn = preproc.preprocess_train(x_train, y_train, verbose=verbose)
val = preproc.preprocess_test(x_test, y_test, verbose=verbose)
if not preproc.get_classes() and verbose:
print('task: text regression (supply class_names argument if this is supposed to be classification task)')
else:
print('task: text classification')
return (trn, val, preproc)</code></pre>
</details>
</dd>
<dt id="ktrain.text.data.texts_from_csv"><code class="name flex">
<span>def <span class="ident">texts_from_csv</span></span>(<span>train_filepath, text_column, label_columns=[], val_filepath=None, max_features=20000, maxlen=400, val_pct=0.1, ngram_range=1, preprocess_mode='standard', encoding=None, lang=None, sep=',', is_regression=False, random_state=None, verbose=1)</span>
</code></dt>
<dd>
<div class="desc"><pre><code>Loads text data from CSV or TSV file. Class labels are assumed to be
one of the following formats:
1. one-hot-encoded or multi-hot-encoded arrays representing classes:
Example with label_columns=['positive', 'negative'] and text_column='text':
text|positive|negative
I like this movie.|1|0
I hated this movie.|0|1
Classification will have a single one in each row: [[1,0,0], [0,1,0]]]
Multi-label classification will have one more ones in each row: [[1,1,0], [0,1,1]]
2. labels are in a single column of string or integer values representing classs labels
Example with label_columns=['label'] and text_column='text':
text|label
I like this movie.|positive
I hated this movie.|negative
3. labels are a single column of numerical values for text regression
NOTE: Must supply is_regression=True for labels to be treated as numerical targets
wine_description|wine_price
Exquisite wine!|100
Wine for budget shoppers|8
Args:
train_filepath(str): file path to training CSV
text_column(str): name of column containing the text
label_column(list): list of columns that are to be treated as labels
val_filepath(string): file path to test CSV. If not supplied,
10% of documents in training CSV will be
used for testing/validation.
max_features(int): max num of words to consider in vocabulary
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if val_filepath is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
encoding (str): character encoding to use. Auto-detected if None
lang (str): language. Auto-detected if None.
sep(str): delimiter for CSV (comma is default)
is_regression(bool): If True, integer targets will be treated as numerical targets instead of class IDs
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random
verbose (boolean): verbosity
</code></pre></div>
<details class="source">
<summary>
<span>Expand source code</span>
</summary>
<pre><code class="python">def texts_from_csv(train_filepath,
text_column,
label_columns = [],
val_filepath=None,
max_features=MAX_FEATURES, maxlen=MAXLEN,
val_pct=0.1, ngram_range=1, preprocess_mode='standard',
encoding=None, # auto-detected
lang=None, # auto-detected
sep=',',
is_regression=False,
random_state=None,
verbose=1):
"""
```
Loads text data from CSV or TSV file. Class labels are assumed to be
one of the following formats:
1. one-hot-encoded or multi-hot-encoded arrays representing classes:
Example with label_columns=['positive', 'negative'] and text_column='text':
text|positive|negative
I like this movie.|1|0
I hated this movie.|0|1
Classification will have a single one in each row: [[1,0,0], [0,1,0]]]
Multi-label classification will have one more ones in each row: [[1,1,0], [0,1,1]]
2. labels are in a single column of string or integer values representing classs labels
Example with label_columns=['label'] and text_column='text':
text|label
I like this movie.|positive
I hated this movie.|negative
3. labels are a single column of numerical values for text regression
NOTE: Must supply is_regression=True for labels to be treated as numerical targets
wine_description|wine_price
Exquisite wine!|100
Wine for budget shoppers|8
Args:
train_filepath(str): file path to training CSV
text_column(str): name of column containing the text
label_column(list): list of columns that are to be treated as labels
val_filepath(string): file path to test CSV. If not supplied,
10% of documents in training CSV will be
used for testing/validation.
max_features(int): max num of words to consider in vocabulary
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if val_filepath is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
encoding (str): character encoding to use. Auto-detected if None
lang (str): language. Auto-detected if None.
sep(str): delimiter for CSV (comma is default)
is_regression(bool): If True, integer targets will be treated as numerical targets instead of class IDs
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random
verbose (boolean): verbosity
```
"""
if encoding is None:
with open(train_filepath, 'rb') as f:
#encoding = chardet.detect(f.read())['encoding']
#encoding = 'utf-8' if encoding.lower() in ['ascii', 'utf8', 'utf-8'] else encoding
encoding = TU.detect_encoding(f.read())
U.vprint('detected encoding: %s (if wrong, set manually)' % (encoding), verbose=verbose)
train_df = pd.read_csv(train_filepath, encoding=encoding,sep=sep)
val_df = pd.read_csv(val_filepath, encoding=encoding,sep=sep) if val_filepath is not None else None
return texts_from_df(train_df,
text_column,
label_columns=label_columns,
val_df = val_df,
max_features=max_features,
maxlen=maxlen,
val_pct=val_pct,
ngram_range=ngram_range,
preprocess_mode=preprocess_mode,
lang=lang, is_regression=is_regression, random_state=random_state,
verbose=verbose)</code></pre>
</details>
</dd>
<dt id="ktrain.text.data.texts_from_df"><code class="name flex">
<span>def <span class="ident">texts_from_df</span></span>(<span>train_df, text_column, label_columns=[], val_df=None, max_features=20000, maxlen=400, val_pct=0.1, ngram_range=1, preprocess_mode='standard', lang=None, is_regression=False, random_state=None, verbose=1)</span>
</code></dt>
<dd>
<div class="desc"><pre><code>Loads text data from Pandas dataframe file. Class labels are assumed to be
one of the following formats:
1. one-hot-encoded or multi-hot-encoded arrays representing classes:
Example with label_columns=['positive', 'negative'] and text_column='text':
text|positive|negative
I like this movie.|1|0
I hated this movie.|0|1
Classification will have a single one in each row: [[1,0,0], [0,1,0]]]
Multi-label classification will have one more ones in each row: [[1,1,0], [0,1,1]]
2. labels are in a single column of string or integer values representing class labels
Example with label_columns=['label'] and text_column='text':
text|label
I like this movie.|positive
I hated this movie.|negative
3. labels are a single column of numerical values for text regression
NOTE: Must supply is_regression=True for integer labels to be treated as numerical targets
wine_description|wine_price
Exquisite wine!|100
Wine for budget shoppers|8
Args:
train_df(dataframe): Pandas dataframe
text_column(str): name of column containing the text
label_columns(list): list of columns that are to be treated as labels
val_df(dataframe): file path to test dataframe. If not supplied,
10% of documents in training df will be
used for testing/validation.
max_features(int): max num of words to consider in vocabulary.
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if val_filepath is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
lang (str): language. Auto-detected if None.
is_regression(bool): If True, integer targets will be treated as numerical targets instead of class IDs
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random
verbose (boolean): verbosity
</code></pre></div>
<details class="source">
<summary>
<span>Expand source code</span>
</summary>
<pre><code class="python">def texts_from_df(train_df,
text_column,
label_columns = [],
val_df=None,
max_features=MAX_FEATURES, maxlen=MAXLEN,
val_pct=0.1, ngram_range=1, preprocess_mode='standard',
lang=None, # auto-detected
is_regression=False,
random_state=None,
verbose=1):
"""
```
Loads text data from Pandas dataframe file. Class labels are assumed to be
one of the following formats:
1. one-hot-encoded or multi-hot-encoded arrays representing classes:
Example with label_columns=['positive', 'negative'] and text_column='text':
text|positive|negative
I like this movie.|1|0
I hated this movie.|0|1
Classification will have a single one in each row: [[1,0,0], [0,1,0]]]
Multi-label classification will have one more ones in each row: [[1,1,0], [0,1,1]]
2. labels are in a single column of string or integer values representing class labels
Example with label_columns=['label'] and text_column='text':
text|label
I like this movie.|positive
I hated this movie.|negative
3. labels are a single column of numerical values for text regression
NOTE: Must supply is_regression=True for integer labels to be treated as numerical targets
wine_description|wine_price
Exquisite wine!|100
Wine for budget shoppers|8
Args:
train_df(dataframe): Pandas dataframe
text_column(str): name of column containing the text
label_columns(list): list of columns that are to be treated as labels
val_df(dataframe): file path to test dataframe. If not supplied,
10% of documents in training df will be
used for testing/validation.
max_features(int): max num of words to consider in vocabulary.
Note: This is only used for preprocess_mode='standard'.
maxlen(int): each document can be of most <maxlen> words. 0 is used as padding ID.
ngram_range(int): size of multi-word phrases to consider
e.g., 2 will consider both 1-word phrases and 2-word phrases
limited by max_features
val_pct(float): Proportion of training to use for validation.
Has no effect if val_filepath is supplied.
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
lang (str): language. Auto-detected if None.
is_regression(bool): If True, integer targets will be treated as numerical targets instead of class IDs
random_state(int): If integer is supplied, train/test split is reproducible.
If None, train/test split will be random
verbose (boolean): verbosity
```
"""
# read in train and test data
train_df = train_df.copy()
train_df[text_column].fillna('fillna', inplace=True)
if val_df is not None:
val_df = val_df.copy()
val_df[text_column].fillna('fillna', inplace=True)
else:
train_df, val_df = train_test_split(train_df, test_size=val_pct, random_state=random_state)
# transform labels
ytransdf = U.YTransformDataFrame(label_columns, is_regression=is_regression)
t_df = ytransdf.apply_train(train_df)
v_df = ytransdf.apply_test(val_df)
class_names = ytransdf.get_classes()
new_lab_cols = ytransdf.get_label_columns(squeeze=True)
x_train = t_df[text_column].values
y_train = t_df[new_lab_cols].values
x_test = v_df[text_column].values
y_test = v_df[new_lab_cols].values
# detect language
if lang is None: lang = TU.detect_lang(x_train)
check_unsupported_lang(lang, preprocess_mode)
# return preprocessed the texts
preproc_type = tpp.TEXT_PREPROCESSORS.get(preprocess_mode, None)
if None: raise ValueError('unsupported preprocess_mode')
preproc = preproc_type(maxlen,
max_features,
class_names = class_names,
lang=lang, ngram_range=ngram_range)
trn = preproc.preprocess_train(x_train, y_train, verbose=verbose)
val = preproc.preprocess_test(x_test, y_test, verbose=verbose)
# QUICKFIX for #314
preproc.ytransform.le = ytransdf.le
return (trn, val, preproc)</code></pre>
</details>
</dd>
<dt id="ktrain.text.data.texts_from_folder"><code class="name flex">
<span>def <span class="ident">texts_from_folder</span></span>(<span>datadir, classes=None, max_features=20000, maxlen=400, ngram_range=1, train_test_names=['train', 'test'], preprocess_mode='standard', encoding=None, lang=None, val_pct=0.1, random_state=None, verbose=1)</span>
</code></dt>
<dd>
<div class="desc"><pre><code>Returns corpus as sequence of word IDs.
Assumes corpus is in the following folder structure:
├── datadir
│ ├── train
│ │ ├── class0 # folder containing documents of class 0
│ │ ├── class1 # folder containing documents of class 1
│ │ ├── class2 # folder containing documents of class 2
│ │ └── classN # folder containing documents of class N
│ └── test
│ ├── class0 # folder containing documents of class 0
│ ├── class1 # folder containing documents of class 1
│ ├── class2 # folder containing documents of class 2
│ └── classN # folder containing documents of class N
Each subfolder should contain documents in plain text format.
If train and test contain additional subfolders that do not represent
classes, they can be ignored by explicitly listing the subfolders of
interest using the classes argument.
Args:
datadir (str): path to folder
classes (list): list of classes (subfolders to consider).
This is simply supplied as the categories argument
to sklearn's load_files function.
max_features (int): maximum number of unigrams to consider
Note: This is only used for preprocess_mode='standard'.
maxlen (int): maximum length of tokens in document
ngram_range (int): If > 1, will include 2=bigrams, 3=trigrams and bigrams
train_test_names (list): list of strings represnting the subfolder
name for train and validation sets
if test name is missing, <val_pct> of training
will be used for validation
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
encoding (str): character encoding to use. Auto-detected if None
lang (str): language. Auto-detected if None.
val_pct(float): Onlyl used if train_test_names has 1 and not 2 names
random_state(int): If integer is supplied, train/test split is reproducible.
IF None, train/test split will be random
verbose (bool): verbosity
</code></pre></div>
<details class="source">
<summary>
<span>Expand source code</span>
</summary>
<pre><code class="python">def texts_from_folder(datadir, classes=None,
max_features=MAX_FEATURES, maxlen=MAXLEN,
ngram_range=1,
train_test_names=['train', 'test'],
preprocess_mode='standard',
encoding=None, # detected automatically
lang=None, # detected automatically
val_pct=0.1, random_state=None,
verbose=1):
"""
```
Returns corpus as sequence of word IDs.
Assumes corpus is in the following folder structure:
├── datadir
│ ├── train
│ │ ├── class0 # folder containing documents of class 0
│ │ ├── class1 # folder containing documents of class 1
│ │ ├── class2 # folder containing documents of class 2
│ │ └── classN # folder containing documents of class N
│ └── test
│ ├── class0 # folder containing documents of class 0
│ ├── class1 # folder containing documents of class 1
│ ├── class2 # folder containing documents of class 2
│ └── classN # folder containing documents of class N
Each subfolder should contain documents in plain text format.
If train and test contain additional subfolders that do not represent
classes, they can be ignored by explicitly listing the subfolders of
interest using the classes argument.
Args:
datadir (str): path to folder
classes (list): list of classes (subfolders to consider).
This is simply supplied as the categories argument
to sklearn's load_files function.
max_features (int): maximum number of unigrams to consider
Note: This is only used for preprocess_mode='standard'.
maxlen (int): maximum length of tokens in document
ngram_range (int): If > 1, will include 2=bigrams, 3=trigrams and bigrams
train_test_names (list): list of strings represnting the subfolder
name for train and validation sets
if test name is missing, <val_pct> of training
will be used for validation
preprocess_mode (str): Either 'standard' (normal tokenization) or one of {'bert', 'distilbert'}
tokenization and preprocessing for use with
BERT/DistilBert text classification model.
encoding (str): character encoding to use. Auto-detected if None
lang (str): language. Auto-detected if None.
val_pct(float): Onlyl used if train_test_names has 1 and not 2 names
random_state(int): If integer is supplied, train/test split is reproducible.
IF None, train/test split will be random
verbose (bool): verbosity
```
"""
# check train_test_names
if len(train_test_names) < 1 or len(train_test_names) > 2:
raise ValueError('train_test_names must have 1 or two elements for train and optionally validation')
# read in training and test corpora
train_str = train_test_names[0]
train_b = load_files(os.path.join(datadir, train_str), shuffle=True, categories=classes)
if len(train_test_names) > 1:
test_str = train_test_names[1]
test_b = load_files(os.path.join(datadir, test_str), shuffle=False, categories=classes)
x_train = train_b.data
y_train = train_b.target
x_test = test_b.data
y_test = test_b.target
else:
x_train, x_test, y_train, y_test = train_test_split(train_b.data,
train_b.target,
test_size=val_pct,
random_state=random_state)
# decode based on supplied encoding
if encoding is None:
encoding = TU.detect_encoding(x_train)
U.vprint('detected encoding: %s' % (encoding), verbose=verbose)