-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathclan.py
1005 lines (741 loc) · 43.6 KB
/
clan.py
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
from multiprocessing import freeze_support
import pylangacq
import json
import mlconjug3
import spacy
# Run this in terminal for Spanish processing
# python -m spacy download es_core_news_md
if __name__ == '__main__':
freeze_support()
###### QUICKSTART TESTING CODE ######
#data = "*CHI:\tI want cookie .\n*MOT:\tokay ."
#reader = pylangacq.Reader.from_strs([data])
#len(reader.utterances())
#reader.n_files()
#reader.utterances()
#reader = pylangacq.Reader.from_strs([eve])
english_conjugator = mlconjug3.Conjugator(language='en')
spanish_conjugator = mlconjug3.Conjugator(language='es')
####### IMPORTING DATA FROM CHILDES ######
def import_english_corpora():
child_urls = {"https://childes.talkbank.org/data/Eng-NA/Brown.zip", "https://childes.talkbank.org/data/Eng-NA/Bloom.zip",
"https://childes.talkbank.org/data/Clinical-MOR/Gillam.zip", "https://childes.talkbank.org/data/Eng-NA/Bohannon.zip",
"https://childes.talkbank.org/data/Eng-NA/Gathercole.zip", "https://childes.talkbank.org/data/Eng-NA/Haggerty.zip",
"https://childes.talkbank.org/data/Eng-NA/Garvey.zip", "https://childes.talkbank.org/data/Eng-NA/Braunwald.zip",
"https://childes.talkbank.org/data/Eng-NA/Warren.zip", "https://childes.talkbank.org/data/Eng-NA/Brent.zip",
"https://childes.talkbank.org/access/Eng-NA/Weist.html", "https://childes.talkbank.org/data/Eng-NA/Clark.zip",
"https://childes.talkbank.org/data/Eng-UK/Fletcher.zip", "https://childes.talkbank.org/data/Eng-NA/Demetras1.zip",
"https://childes.talkbank.org/data/Eng-UK/Wells.zip", "https://childes.talkbank.org/data/Eng-NA/Demetras2.zip",
"https://childes.talkbank.org/data/Eng-UK/Tommerdahl.zip", "https://childes.talkbank.org/data/Clinical-MOR/EllisWeismer.zip",
"https://childes.talkbank.org/data/Clinical-MOR/Conti/Conti1.zip", "https://childes.talkbank.org/data/Eng-NA/Evans.zip",
"https://childes.talkbank.org/data/Eng-NA/Bates.zip", "https://childes.talkbank.org/data/Eng-NA/Feldman.zip",
"https://childes.talkbank.org/data/Eng-NA/Bernstein.zip", "https://childes.talkbank.org/data/Eng-NA/Hall.zip",
"https://childes.talkbank.org/data/Eng-NA/Bliss.zip", "https://childes.talkbank.org/data/Eng-NA/Higginson.zip",
"https://childes.talkbank.org/data/Eng-NA/HSLLD.zip", "https://childes.talkbank.org/data/Eng-NA/Kuczaj.zip",
"https://childes.talkbank.org/data/Eng-NA/McCune.zip", "https://childes.talkbank.org/data/Eng-NA/Morisset.zip",
"https://asd.talkbank.org/data/English/Nadig.zip", "https://childes.talkbank.org/data/Eng-NA/Nelson.zip",
"https://childes.talkbank.org/data/Eng-NA/NewEngland.zip", "https://childes.talkbank.org/data/Eng-NA/NewmanRatner.zip",
"https://childes.talkbank.org/data/Clinical-MOR/Nicholas/TD.zip", "https://childes.talkbank.org/data/Eng-NA/Peters.zip",
"https://childes.talkbank.org/data/Clinical-MOR/POLER.zip", "https://childes.talkbank.org/data/Eng-NA/Post.zip",
"https://childes.talkbank.org/data/Eng-NA/Rollins.zip", "https://childes.talkbank.org/data/Clinical-MOR/Rondal/TD.zip",
"https://childes.talkbank.org/data/Eng-NA/Sachs.zip", "https://childes.talkbank.org/data/Eng-NA/Sawyer.zip",
"https://childes.talkbank.org/data/Eng-NA/Snow.zip", "https://childes.talkbank.org/data/Eng-NA/Soderstrom.zip",
"https://childes.talkbank.org/data/Eng-NA/Sprott.zip", "https://childes.talkbank.org/data/Eng-NA/Suppes.zip",
"https://childes.talkbank.org/data/Eng-NA/Tardif.zip", "https://childes.talkbank.org/data/Eng-NA/Valian.zip",
"https://childes.talkbank.org/data/Eng-NA/VanHouten.zip", "https://childes.talkbank.org/data/Eng-NA/VanKleeck.zip",
"https://childes.talkbank.org/data/Eng-UK/Belfast.zip", "https://childes.talkbank.org/data/Eng-UK/Edinburgh.zip",
"https://phon.talkbank.org/data/Eng-UK/Cruttenden.zip", "https://childes.talkbank.org/data/Eng-UK/Forrester.zip",
"https://childes.talkbank.org/data/Eng-UK/Howe.zip", "https://childes.talkbank.org/data/Eng-UK/KellyQuigley.zip",
"https://childes.talkbank.org/data/Eng-UK/Korman.zip", "https://childes.talkbank.org/data/Eng-UK/Lara.zip",
"https://childes.talkbank.org/data/Eng-UK/Manchester.zip", "https://childes.talkbank.org/data/Eng-UK/Nuffield.zip",
"https://childes.talkbank.org/data/Eng-UK/OdiaMAIN.zip", "https://asd.talkbank.org/data/English/QuigleyMcNally.zip",
"https://childes.talkbank.org/data/Eng-UK/Sekali.zip", "https://phon.talkbank.org/data/Eng-UK/Smith.zip",
"https://childes.talkbank.org/data/Eng-UK/Thomas.zip", "https://childes.talkbank.org/data/Clinical-MOR/ENNI.zip",
"https://childes.talkbank.org/data/Clinical-MOR/UCSD.zip"
}
english_corpora = []
for url in child_urls:
english_corpora.append(pylangacq.read_chat(url))
return english_corpora
def import_biling_child():
# BILINGUAL DATA
perez_url = "https://childes.talkbank.org/data/Biling/Perez.zip"
perez = pylangacq.read_chat(perez_url)
perez_tokens = perez.tokens(participants="CHI")
# not POS notated
montes_url = "https://childes.talkbank.org/data/Spanish/Montes.zip"
montes = pylangacq.read_chat(montes_url)
#1-2yo, pos notated
reyes_url = "https://childes.talkbank.org/data/Biling/Reyes.zip"
reyes = pylangacq.read_chat(reyes_url)
# annotated, not labelled as CHI
deuchar_url = "https://childes.talkbank.org/data/Biling/Deuchar.zip"
deuchar = pylangacq.read_chat(deuchar_url)
silva_url = "https://childes.talkbank.org/data/Biling/SilvaCorvalan.zip"
silva = pylangacq.read_chat(silva_url)
fuertes_url = "https://childes.talkbank.org/data/Biling/FerFuLice.zip"
fuertes = pylangacq.read_chat(fuertes_url)
# POS annotated
biling_child_corpora = [fuertes, silva, deuchar, reyes, montes]
url_list = ["https://childes.talkbank.org/data/Biling/Perez.zip",
"https://childes.talkbank.org/data/Frogs/Spanish-MiamiBiling.zip"
]
unannotated_list = []
for url in url_list:
unannotated_list.append(pylangacq.read_chat(url))
return [biling_child_corpora, unannotated_list]
def import_biling_adult():
# BILINGUAL ADULT DATA
miami_url = "https://biling.talkbank.org/data/Bangor/Miami.zip"
miami = pylangacq.read_chat(miami_url)
miami_tokens = miami.tokens()
return miami_tokens
def import_spanish_child():
# SPANISH MONO CHILD DATA
annotated_urls = ["https://childes.talkbank.org/data/Spanish/Marrero.zip", "https://childes.talkbank.org/data/Spanish/Hess.zip",
"https://phon.talkbank.org/data/Spanish/Koine.zip","https://childes.talkbank.org/data/Frogs/Spanish-Aguilar.zip",
"https://childes.talkbank.org/data/Spanish/FernAguado.zip", "https://childes.talkbank.org/data/Spanish/ColMex.zip",
"https://childes.talkbank.org/data/Spanish/OreaPine.zip","https://childes.talkbank.org/data/Spanish/Aguirre.zip",
"https://childes.talkbank.org/data/Spanish/GRERLI.zip","https://childes.talkbank.org/data/Spanish/Linaza.zip",
"https://childes.talkbank.org/data/Spanish/Ornat.zip","https://childes.talkbank.org/data/Spanish/Remedi.zip",
"https://childes.talkbank.org/data/Spanish/Romero.zip", "https://childes.talkbank.org/data/Spanish/SerraSole.zip"
]
spa_corpora = []
for url in annotated_urls:
spa_corpora.append(pylangacq.read_chat(url))
url_list = ["https://childes.talkbank.org/data/Spanish/BecaCESNo.zip", # 3-11, Child: CHI, Adult: MOT, PAT, AUN, INT, EXP, PAU? (too much overlap)
"https://childes.talkbank.org/data/Spanish/DiezItza.zip", # 3yo, Child: CHI, Adult: MOT, INV
"https://childes.talkbank.org/data/Frogs/Spanish-Ornat.zip", # 3-5, 8, Child: CHI
"https://childes.talkbank.org/data/Spanish/Shiro.zip", # not annotated
"https://childes.talkbank.org/data/Frogs/Spanish-Sebastian.zip", # 3-5, 9, 20, CHI: CHI
"https://childes.talkbank.org/data/Spanish/JacksonThal.zip",
"https://childes.talkbank.org/data/Spanish/Nieva.zip",
"https://childes.talkbank.org/data/Spanish/Vila.zip"]
unannotated_list = []
for url in url_list:
unannotated_list.append(pylangacq.read_chat(url))
return [spa_corpora, unannotated_list]
###
english_cp_verbs = ['say', 'know', 'think', 'believe', 'suggest', 'find', 'show', 'mean', 'see', 'realize',
'note', 'argue', 'indicate', 'feel', 'understand', 'hope', 'report', 'state', 'assume', 'agree',
'claim', 'ensure', 'learn', 'remember', 'conclude', 'hear', 'admit', 'notice', 'decide', 'add',
'insist', 'recognize', 'tell', 'explain', 'discover', 'reveal', 'prove', 'seem', 'acknowledge',
'announce', 'estimate', 'demonstrate', 'write', 'mention', 'imply', 'appear', 'confirm',
'imagine', 'require', 'suspect', 'worry', 'forget', 'expect', 'demand', 'determine',
'recommend', 'complain', 'assert', 'declare', 'fear', 'predict', 'warn', 'deny', 'ask', 'accept',
'doubt', 'observe', 'contend', 'consider', 'maintain', 'guess', 'read', 'wish', 'rule',
'emphasize', 'propose', 'pretend', 'use', 'love', 'recall', 'pray', 'suppose', 'allege', 'testify',
'figure', 'concede', 'stress', 'speculate', 'guarantee', 'call', 'convince',
'promise', 'assure', 'teach']
# Deleted: "let", "make", "hold", "like" (not a bad CP taking verb, null C is too easily mistaken), "happen", "want?"
spanish_cp_verbs = ['decir', 'ser', 'asegurar', 'creer', 'saber', 'contar', 'señalar', 'afirmar', 'pensar',
'recordar', 'indicar', 'explicar', 'querer', 'aclarar', 'comentar', 'esperar', 'ver', 'revelar',
'agregar', 'parecer', 'hacer','manifestar', 'confesar', 'demostrar', 'considerar', 'sostener',
'sentir', 'entender', 'reconocer', 'confirmar', 'declarar', 'descubrir', 'argumentar', 'informar',
'añadir', 'mencionar', 'responder', 'alegar', 'destacar', 'relatar', 'admitir', 'suponer', 'dejar',
'gustar', 'conocer', 'negar', 'expresar', 'evitar', 'anunciar', 'significar', 'permitir', 'enterar',
'imaginar', 'denunciar', 'notar', 'sospechar', 'decidir', 'aceptar', 'determinar', 'precisar',
'resultar', 'aprovechar', 'mostrar', 'recalcar', 'advertir', 'detallar', 'olvidar', 'lograr', 'impedir',
'reiterar', 'provocar', 'referir', 'comprobar', 'pedir', 'resaltar', 'apuntar', 'subrayar', 'sugerir',
'contestar', 'aseverar', 'establecer', 'escuchar', 'insinuar', 'acordar', 'descartar', 'remarcar',
'presumir', 'probar', 'asumir', 'enfatizar', 'narrar', 'desear', 'temer', 'lamentar', 'importar',
'aducir', 'defender', 'conseguir', 'desmentir']
# Deleted: "poder"
### TESTING/EXPLORING FUNCTIONS ###
def print_sample(corpus):
# Prints sample of corpus, change i to specify length of sample
for i , token in enumerate(corpus):
# Print first 10 words
if i < 10:
print("word:", token.word + " | " + "part-of-speech tag:", token.pos + "|" + "morph:", token.mor)
def print_sample_from(corpus,word):
# Prints a sample starting from a particular word (helpful for finding a language in biling corpus)
printing = False
counter = 0
found = False
# Reference for syntax to call these attributes
#print("~~" + english_child_tokens[i-1].word + " " + english_child_tokens[i].word + " " + english_child_tokens[i+1].word)
#print("word:", english_child_tokens[i-1].mor + " | " + "part-of-speech tag:", english_child_tokens[i-1].pos)
#print("word:", token.word + " | " + "part-of-speech tag:", token.pos)
#print("word:", english_child_tokens[i+1].word + " | " + "part-of-speech tag:", english_child_tokens[i+1].pos)
#print("morphological information:", token.mor)
#print("grammatical relation:", token.gra)
for token in corpus:
try:
# Find input word
if token.word == word and found == False:
printing = True
found = True
if printing == True and counter < 10:
try:
print("word:", token.word + " | " + "part-of-speech tag:", token.pos + "|" + "morph:", token.mor)
except TypeError:
print("word:", token.word + " | part-of-speech tag: - | morph: -")
counter += 1
elif counter == 10:
break
except AttributeError:
continue
def print_thats(corpus):
# Prints all instances of "that"
for i , token in enumerate(corpus):
last_token = corpus[i-1]
#last_pos = wn.synsets(last_token.word)[0].pos()
# Print first 60 words
if i > 0 and i < len(corpus)-4 and token.word == "that" and last_token.pos == "v":
last_token = corpus[i-1]
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
print(last_token.word, token.word + " ("+ token.pos +")" , token2.word, token3.word, token4.word)
def print_ques(corpus):
# Prints all instances of "que"
for i , token in enumerate(corpus):
last_token = corpus[i-1]
#last_pos = wn.synsets(last_token.word)[0].pos()
# Print first 60 words
if i > 0 and i < len(corpus)-4 and token.word == "que":
last_token = corpus[i-1]
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
try:
print(last_token.word, token.word + " (" + token.pos + ") ", token2.word, token3.word, token4.word)
except TypeError:
print(last_token.word, token.word + " (-) ", token2.word, token3.word, token4.word)
def print_que_cs(corpus):
# Prints all instances of "que" where "que" is labelled as a complementizer
for i , token in enumerate(corpus):
last_token = corpus[i-1]
matrix_pos = ["cop", "v","inf"]
embedded_pos = ["cop", "v"]
# Find a que
if i > 0 and i < len(corpus)-4 and token.word == "que":
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
try:
last_pos = last_token.pos.split("|")[0]
next_pos = token2.pos.split("|")[0]
third_pos = token3.pos.split("|")[0]
fourth_pos = token4.pos.split("|")[0]
except AttributeError:
continue
embedded_conj_verb = False
# Want to make sure the next verb isn't an infinitive
if next_pos in embedded_pos:
embedded_conj_verb = True
elif next_pos != "inf" and third_pos in embedded_pos:
embedded_conj_verb = True
elif next_pos != "inf" and third_pos != "inf" and fourth_pos in embedded_pos:
embedded_conj_verb = True
if last_pos in matrix_pos and embedded_conj_verb == True:
try:
print(last_token.word, token.word + " (" + token.pos + ") ", token2.word, token3.word, token4.word)
except TypeError:
print(last_token.word, token.word + " (-) ", token2.word, token3.word, token4.word)
def print_null_ques(corpus):
# Prints instances of null "que"
for i , token in enumerate(corpus):
matrix_pos = ["cop", "v","inf"]
embedded_pos = ["cop", "v"]
# Find a que
if i > 0 and i < len(corpus)-4 and token.pos in matrix_pos:
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
try:
next_pos = token2.pos.split("|")[0]
third_pos = token3.pos.split("|")[0]
fourth_pos = token4.pos.split("|")[0]
except AttributeError:
continue
embedded_conj_verb = False
# Want to make sure the next verb isn't an infinitive
if next_pos in embedded_pos:
embedded_conj_verb = True
elif next_pos != "inf" and third_pos in embedded_pos:
embedded_conj_verb = True
elif next_pos != "inf" and third_pos != "inf" and fourth_pos in embedded_pos:
embedded_conj_verb = True
if token.pos in matrix_pos and embedded_conj_verb == True:
try:
print(token.word, token2.word, token3.word, token4.word)
except TypeError:
print(token.word, token2.word, token3.word, token4.word)
def print_conjugations(conjugator, verb):
# This is basically from the mlconjug3 example
test_verb = conjugator.conjugate(verb)
all_conjugated_forms = test_verb.iterate()
print(all_conjugated_forms)
### HELPER FUNCTIONS ###
def expand_cp_verbs(verb_array, conjugator):
cp_conjugations = {}
bad_verbs = ["me", "mí", "sí", "se"]
# these are most commonly used as clitics, but mostrar conjugates to me apparently?
for verb in verb_array:
this_verb = conjugator.conjugate(verb)
conjugated_forms = this_verb.iterate()
cp_conjugations[verb] = [verb]
for form in conjugated_forms:
#print(form)
try:
token = form[3]
except IndexError:
#sometimes for the infinitive the tuples are shorter-- we are contributing inifinitive so don't need it
continue
if token in cp_conjugations[verb] or token in bad_verbs:
continue
## This covers both Spanish and English
elif str(form[0])[0:6] == "indica" or str(form[0])[0:6] == "Indica":
cp_conjugations[verb].append(token)
else:
continue
return cp_conjugations
def is_okay_intermediate(verb, subsequent_tokens):
# In searching for an embedded finite verb, we need to allow space for an NP (especially in English).
# Here we allow for 3 tokens between verbs, which should cover an NP
# This function makes sure that those three (or fewer) intermediates do not contain:
# Sentence endings, reduplication, adjectives as an argument, prepositions, wh words, complementizers, or infinitives
is_okay = True
subsequent_words= []
punct_stops = [ ".", "?", ",", "!", "/", "/.", '"/.', "//.","..."]
spa_stops = ["que", "qué", "o", "y", "pero", "dónde", "cómo", "más", "cuándo", "donde", "cuando", "como", "mas","sí", "si", "quién", "cuánto", "porqué","porque", "de","cual", "cuántas", "cuántos", "pues"]
eng_stops = ["who", "what", "where", "when", "why", "if", "how", "because"]
# Consider adding ahora, as and adjectives after ser
indef_stops = ["para", "así"]
stops = punct_stops + spa_stops + indef_stops + eng_stops
for token in subsequent_tokens:
subsequent_words.append(token.word)
if token.word in stops:
# Checks for sentence endings, prepositions,
is_okay = False
adj_set = ["ADJ", "Adj", "adj"]
if verb in subsequent_words:
# Checks for reduplication
is_okay = False
#elif subsequent_tokens[0].pos in adj_set:
# # Checks for adjective following matrix (particularly important for copula)
# is_okay = False
elif "to" in subsequent_words:
is_okay = False
elif subsequent_tokens[-1].pos == "inf":
is_okay = False
return is_okay
def is_good_matrix_verb(token, verb_conj):
is_good = False
verb_root = "-"
for root in verb_conj:
#print(spanish_cp_conj)
if token in verb_conj[root]:
is_good = True
verb_root = root
return [is_good, verb_root]
return [is_good, verb_root]
def good_embedded_verb(token):
infinitive_endings = ["rlo", "rla", "rle", "rlos", "rlas", "rles","rte", "rme", "rse", "ando", "rnos"]
bad_verbs = ["este", "corta", "corto", "larga", "nada", "completo", "completa", "cumple", "llama", "duro", "tardes", "colores", "//.", "poque", "polque", "corriendo", "daño", "vale", "iguales"]
good_verb = False
emb_verb_types = ["v", "cop", "mod", "aux", "part"]
if token.pos in emb_verb_types:
good_verb = True
if str(token.word)[-1] == "r":
good_verb = False
elif str(token.word)[-3:-1] in infinitive_endings:
good_verb = False
elif str(token.word)[-4:-1] in infinitive_endings:
good_verb = False
elif str(token.word)[-5:-1] in infinitive_endings:
good_verb = False
elif token.word in bad_verbs:
good_verb = False
return good_verb
### INTERMEDIATE STEPS ###
def split_into_ages(corpus_array):
# Takes in array of corpora and returns child utterances separated by age
max_age = 11
token_array = []
i =0
while i < max_age:
token_array.append([])
i +=1
failed_files = 0
sli_files = 0
for corpus in corpus_array:
for i, header in enumerate(corpus.headers()):
try:
# Rule out SLI files
if 'SLI' in header['Types']:
sli_files += 1
continue
else:
transcript = corpus[i]
for participant in header['Participants']:
try:
part_age = header['Participants'][participant]['age'].split(';')[0]
if part_age != "":
int_age = int(part_age)
if int_age >= max_age:
continue
else:
part_tokens = transcript.tokens(participants = participant)
# perhaps add CHI2, I don't remember if they added the age
token_array[int_age] = token_array[int_age] + part_tokens
else:
continue
except TypeError or IndexError or ValueError:
continue
except KeyError or TypeError:
# If there's no age info, just skip this file
failed_files = failed_files +1
continue
print("Total failed files: " + str(failed_files))
print("Total SLI files: " + str(sli_files))
print("3yo:", len(token_array[3]), ", 4yo:", len(token_array[4]), ", 5yo:", len(token_array[5]))
return token_array
def adult_split_into_ages(corpus_array): #TODO Make this less clunky
max_age = 11
token_array = []
i =0
while i < max_age:
token_array.append([])
i +=1
adult_tags = ["MOT", "FAT", "MAD", "PAD", "INV", "ADU"]
for corpus in corpus_array:
for transcript in corpus:
corpus_tokens = []
for participant in adult_tags:
adult_tokens = transcript.tokens(participants=participant)
corpus_tokens += adult_tokens
try:
age = transcript.ages(months= True)[0]//12
int_age = int(age)
if int_age >= max_age:
continue
else:
token_array[int_age] = token_array[int_age] + corpus_tokens
except TypeError or IndexError or ValueError:
continue
return token_array
def annotate_corpus(corpus_array, participant = "child"):
if participant == "child":
age_corpora = split_into_ages(corpus_array)
elif participant == "adult":
age_corpora = adult_split_into_ages(corpus_array)
nlp = spacy.load('es_core_news_md')
new_token_array = []
corpus_counter = 0
for corpus in age_corpora:
# STEP 1: CONVERT TOKENS TO STRING AND RUN SPACY
filestring = ""
for token in corpus:
filestring = filestring + token.word + " "
document = nlp(filestring)
converted_tokens = []
# STEP 2: CONVERT BACK TO CHAT FORMAT
for token in document:
# I want the pos tagging to be the same as CHAT format
pos = token.pos_
infinitive_endings = ["rlo", "rla", "rle", "rlos", "rlas", "rles","rte", "rme", "rse", "ando"]
if token.pos_ == "VERB":
if str(token.text)[-1] == "r":
pos = "inf"
elif str(token.text)[-3:-1] in infinitive_endings:
pos = "inf"
elif str(token.text)[-4:-1] in infinitive_endings:
pos = "inf"
elif str(token.text)[-5:-1] in infinitive_endings:
pos = "prog"
else:
pos = "v"
newToken = pylangacq.objects.Token(token.text, pos, token.tag_, (0,0,token.dep_))
converted_tokens.append(newToken)
#print_sample(converted_tokens)
new_token_array.append(converted_tokens)
corpus_counter += 1
return new_token_array
def that_counter(corpus):
# Counts instances of null and overt "that" in given corpus and returns two dictionaries with those values
verb_dict = {}
that_counter = 0
emb_verb_types = ["v", "cop", "mod", "aux", "part"]
### STEP 1: Search for instances of that [VERB that]
for i , token in enumerate(corpus):
if i > 0 and i < len(corpus)-4:
last_token = corpus[i-1]
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
token5 = corpus[i+4]
[good_matrix, root] = is_good_matrix_verb(last_token.word,english_cp_conj)
# Check for sequence VERB that (NP) finiteV
if token.word == "that" and last_token.pos == "v" and good_matrix:
embedded_conj_verb = False
if is_okay_intermediate(last_token.word, [token2]) and token2.pos in emb_verb_types:
embedded_conj_verb = True
elif is_okay_intermediate(last_token.word, [token2, token3]) and token3.pos in emb_verb_types:
embedded_conj_verb = True
elif is_okay_intermediate(last_token.word, [token2, token3, token4]) and token4.pos in emb_verb_types:
embedded_conj_verb = True
elif is_okay_intermediate(last_token.word, [token2, token3, token4, token5]) and token5.pos in emb_verb_types:
embedded_conj_verb = True
if embedded_conj_verb == True:
that_counter += 1
# Update frequency in verb dictionary
if root in verb_dict:
verb_dict[root] += 1
else:
verb_dict[root] = 1
# An earlier version pulled the root from the MOR tier of the token, rather than the conjugated verb dictionary
#verb_str = last_token.mor
#verb_array = re.split(r'-|&', verb_str)
#verb = verb_array[0]
print(verb_dict)
### STEP 2: Search for instances of null CP with these same verbs
null_counter = 0
null_dict = {}
null_string = ""
for i, token in enumerate(corpus):
if i < len(corpus) - 4:
tokenminus3 = corpus[i-3]
tokenminus2 = corpus[i-2]
last_token = corpus[i-1]
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
token5 = corpus[i+4]
[good_matrix, root] = is_good_matrix_verb(token.word, english_cp_conj)
# Check that verb is in cp-taking verb list
if good_matrix and token.pos == "v":
embedded_conj_verb = False
if is_okay_intermediate(token.word, [token2]) and token2.pos in emb_verb_types:
embedded_conj_verb = True
elif is_okay_intermediate(token.word, [token2, token3]) and token3.pos in emb_verb_types:
embedded_conj_verb = True
elif is_okay_intermediate(token.word, [token2, token3, token4]) and token4.pos in emb_verb_types:
embedded_conj_verb = True
elif is_okay_intermediate(token.word, [token2, token3, token4, token5]) and token5.pos in emb_verb_types:
embedded_conj_verb = True
if embedded_conj_verb:
null_counter += 1
try:
if root != "say":
try:
this_ex = tokenminus3.word + " " + tokenminus2.word + " " + last_token.word + " " + token.word + " (" + root + ") " + token2.word + " (" + token2.pos + ") " + token3.word + " (" + token3.pos + ") "+ token4.word + " (" +token4.pos +")"
#print(tokenminus2.word, last_token.word, token.word + " (" + root + ") " + token2.word + " (" + token2.pos + ") " + token3.word + " (" + token3.pos + ") "+ token4.word + " (" +token4.pos +")")
null_string = null_string + this_ex + "\n"
except UnicodeEncodeError:
null_string = null_string
except TypeError:
null_string = null_string
# Update frequency in null dictionary
if root in null_dict:
null_dict[root] += 1
else:
null_dict[root] = 1
print(null_dict)
print("'That' appeared as a complementizer " + str(that_counter) + " times")
print("'That' was unpronounced " + str(null_counter) + " times")
this_analysis = [[verb_dict, null_dict], null_string]
return this_analysis
def que_counter(corpus):
# Counts instances of null and overt "que" in given corpus and returns two dictionaries with those values
verb_dict = {}
que_counter = 0
ques = ["que", "qué" ]
bad_predecessors = ["el", "este", "si", "que"]
### STEP 1: Search for instances of that [VERB que ...VERB]
## Excluding infinitives
for i , token in enumerate(corpus):
last_token = corpus[i-1]
# Find a que
[good_matrix, root] = is_good_matrix_verb(last_token.word, spanish_cp_conj)
if i > 0 and i < len(corpus)-4 and token.word in ques and good_matrix:
tokenminus2 = corpus[i-2]
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
token5 = corpus[i+4]
embedded_conj_verb = False
# Want to make sure the next verb isn't an infinitive
if tokenminus2.word in bad_predecessors:
embedded_conj_verb = False
elif is_okay_intermediate(last_token.word, [token2]) and good_embedded_verb(token2):
embedded_conj_verb = True
elif is_okay_intermediate(last_token.word, [token2, token3]) and good_embedded_verb(token3):
embedded_conj_verb = True
elif is_okay_intermediate(last_token.word, [token2, token3, token4]) and good_embedded_verb(token4):
embedded_conj_verb = True
elif is_okay_intermediate(last_token.word, [token2, token3, token4, token5]) and good_embedded_verb(token5):
embedded_conj_verb = True
if embedded_conj_verb:
que_counter += 1
# Update frequency in verb dictionary
if root in verb_dict:
verb_dict[root] += 1
else:
verb_dict[root] = 1
print(verb_dict)
### STEP 2: Search for instances of null CP with these same verbs
null_counter = 0
null_dict = {}
null_string = ""
for i, token in enumerate(corpus):
[good_matrix, root] = is_good_matrix_verb(token.word, spanish_cp_conj)
#print(token.word + root)
if i < len(corpus) - 4 and good_matrix:
tokenminus3 = corpus[i-3]
tokenminus2 = corpus[i-2]
last_token = corpus[i-1]
token2 = corpus[i+1]
token3 = corpus[i+2]
token4 = corpus[i+3]
token5 = corpus[i+4]
embedded_conj_verb = False
# Want to make sure the next verb isn't an infinitive
if last_token.word == "que" or last_token.word in bad_predecessors:
embedded_conj_verb= False
elif is_okay_intermediate(token.word, [token2]) and good_embedded_verb(token2):
embedded_conj_verb = True
elif is_okay_intermediate(token.word, [token2, token3]) and good_embedded_verb(token3):
embedded_conj_verb = True
elif is_okay_intermediate(token.word, [token2, token3, token4]) and good_embedded_verb(token4):
embedded_conj_verb = True
elif is_okay_intermediate(token.word, [token2, token3, token4, token5]) and good_embedded_verb(token5):
embedded_conj_verb = True
if embedded_conj_verb == True:
null_counter += 1
if root != "decir":
try:
this_ex = tokenminus3.word + " " + tokenminus2.word + " " + last_token.word + " " + token.word + " (" + root + ") " + token2.word + " (" + token2.pos + ") " + token3.word + " (" + token3.pos + ") "+ token4.word + " (" +token4.pos +")" + token5.word + " (" +token5.pos +")"
#print(tokenminus2.word, last_token.word, token.word + " (" + root + ") " + token2.word + " (" + token2.pos + ") " + token3.word + " (" + token3.pos + ") "+ token4.word + " (" +token4.pos +")")
except TypeError:
this_ex = ""
null_string = null_string + this_ex + "\n"
# Update frequency in verb dictionary
if root in null_dict:
null_dict[root] += 1
else:
null_dict[root] = 1
print(null_dict)
print("'Que' appeared as a complementizer " + str(que_counter) + " times")
print("'Que' was unpronounced " + str(null_counter) + " times")
this_analysis = [[verb_dict, null_dict], null_string]
return this_analysis
def combine_age_arrays(age_arrays):
#age_arrays should come in as [array1[age1, age2, age3...], array2[age1,age2,age3]]
total_tokens = []
i = 0
while i < len(age_arrays[0]):
total_tokens.append([])
i += 1
for sub_array in age_arrays:
for i, age in enumerate(sub_array):
total_tokens[i] = total_tokens[i] + age
return total_tokens
def dict_to_csv(cp_verbs, dictionary_array, complementizer, label):
# Takes array of paired dictionaries and prints them into a summary csv
filename = label +" "+ complementizer +" frequency_array.csv"
group_counter = 0
# Dictionary array will come in as:
# dictionary_array[age_group1[verb_dict1, null_dict1], age_group[verb_dict2, null_dict1], etc]
with open(filename,"w") as outfile:
## Write the first line of the csv so we know what we're looking at
first_line = " ,"
for age_group in dictionary_array:
first_line = first_line + "group"+ str(group_counter) + " overt, group" + str(group_counter) + " null, "
group_counter = group_counter + 1
first_line = first_line + "\n"
outfile.write(first_line)
## Add information to these columns
for cp_verb in cp_verbs:
# First column
verb_line = cp_verb
for age_group in dictionary_array:
overt_col = 0
null_col = 0
if cp_verb in age_group[0]:
overt_col = age_group[0][cp_verb]
if cp_verb in age_group[1]:
null_col = age_group[1][cp_verb]
verb_line = verb_line +","+ str(overt_col) +","+ str(null_col)
verb_line = verb_line + "\n"
#outfile.write(f"{cp_verb},{col2},{col3}\n")
outfile.write(verb_line)
### PRIMARY FUNCTIONS ###
def that_analysis(age_array, name):
counter = 0
frequency_array = []
if len(age_array) > 1:
counter = 0
filename = name + " instances of null that.txt"
with open(filename,"w") as outfile:
for corpus in age_array:
age_marker = str(counter) + "yo--- \n"
print(age_marker)
[that_counts,exception_string] = that_counter(corpus)
frequency_array.append(that_counts)
#print(exception_string)
#outfile.write(age_marker + "\n")
#outfile.write(exception_string)
counter = counter +1
return dict_to_csv(english_cp_verbs, frequency_array, "that", name)
def que_analysis(age_array, name):
## If not splitting, submit list of tokens to corpus
## If splitting, submit array of corpora
## This is b/c the corpora contain age info that we still need
## TODO: Make this not a problem haha
counter = 0
frequency_array = []
if len(age_array) > 1:
counter = 0
filename = name + " instances of null que.txt"
with open(filename,"w") as outfile:
for corpus in age_array:
age_marker = str(counter) + "yo---"
print(age_marker)
[que_counts, exception_string] = que_counter(corpus)
frequency_array.append(que_counts)
print(exception_string)
outfile.write(age_marker + "\n")
outfile.write(exception_string)
counter = counter +1
return dict_to_csv(spanish_cp_verbs, frequency_array, "que", name)
### MAIN ####
with open("spa_conj", 'r') as f:
spanish_cp_conj = json.load(f)
with open("eng_conj", 'r') as f:
english_cp_conj = json.load(f)
#spanish_cp_conj = expand_cp_verbs(spanish_cp_verbs, spanish_conjugator)
#with open("spa_conj", 'w') as f:
# f.write(json.dumps(spanish_cp_conj, indent=4))
english_corpora = import_english_corpora()
english_ages = split_into_ages(english_corpora)
english_child_dir = adult_split_into_ages(english_corpora)
that_analysis(english_ages, "child")
that_analysis(english_child_dir, "child dir")
# specific setup so we're not loading everything every time
#spanish_cp_conj = expand_cp_verbs(spanish_cp_verbs, spanish_conjugator)
#[spanish_child_corpora, beca] = import_spanish_child()
#biling_adult = [import_biling_adult()]
#que_analysis(biling_adult, "biling adult")
#that_analysis(biling_adult, "biling adult")
#biling_annotated = annotate_corpus(unannotated_biling, "adult")
##biling_ages = adult_split_into_ages(biling_child_corpora)
#total_biling_child = combine_age_arrays([biling_ages, biling_annotated])
#que_analysis(total_biling_child, "total biling dir")
#that_analysis(total_biling_child, "total biling dir")
# Child data
#spa_annotated = annotate_corpus(beca)
#spanish_ages = split_into_ages(spanish_child_corpora)
#total_spanish_child = combine_age_arrays([spa_annotated, spanish_ages])
#que_analysis(total_spanish_child, "spanish child")
#spa_adult_annotated = annotate_corpus(beca, "adult")
#adult_spanish_ages = adult_split_into_ages(spanish_child_corpora)
#total_spanish_child_dir = combine_age_arrays([spa_adult_annotated, adult_spanish_ages])
#que_analysis(total_spanish_child_dir, "spanish child dir")