-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathdata_preprocessing.py
1113 lines (1040 loc) · 37.6 KB
/
data_preprocessing.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
"""
真实数据集预处理
"""
import pickle
import datetime
from typing import Dict
from pathlib import Path
from tqdm.auto import tqdm
import numpy as np
import pandas as pd
from pandas import Series, DataFrame
import networkx as nx
from utils.common import timeit, colored_print, download_file_from_ftp, unzip_file
from utils.toolbox import REAL_DATA_PATH, GraphTool
class DoubanPreprocessor:
RAW_FILENAME = "DoubanDataset-raw.zip"
def __init__(
self,
year_range: tuple[int, int] = (2000, 2020),
root_path: Path = REAL_DATA_PATH / "douban",
):
self.year_range = year_range
self.root_path = root_path
self.raw_path = self.root_path / "raw"
self.processed_path = self.root_path / "processed"
self.movie_data_path = self.raw_path / "movie.csv"
self.person_data_path = self.raw_path / "person.csv"
self.root_path.mkdir(exist_ok=True)
self.raw_path.mkdir(exist_ok=True)
self.processed_path.mkdir(exist_ok=True)
"""原始数据下载"""
if self.movie_data_path.exists() and self.person_data_path.exists():
pass
else:
colored_print("原始数据下载中...")
download_file_from_ftp(self.raw_path, self.RAW_FILENAME)
unzip_file(str(self.raw_path), str(self.raw_path / self.RAW_FILENAME))
colored_print("原始数据已就绪")
"""数据预处理"""
# NOTE: 通过data.pkl是否存在判断是否要重新预处理
if (self.processed_path / Path("data.pkl")).exists():
pass
else:
colored_print("启动数据预处理...")
self.__initialize()
colored_print("数据预处理完成")
def read_processed_data(self):
path = self.processed_path / Path("data.pkl")
with open(path, "rb") as f:
data = pickle.load(f)
return data
def __handle_genre(self, genreAlias):
genreMap = {
"家庭": ["家庭"],
"爱情": ["爱情", "愛情 Romance"],
"歌舞": ["歌舞"],
"古装": ["古装"],
"音乐": ["音乐"],
"同性": ["同性"],
"传记": ["传记", "傳記 Biography"],
"情色": ["情色", "Adult"],
"恐怖": ["惊栗", "惊悚", "驚悚 Thriller", "恐怖", "鬼怪"],
"武侠": ["武侠"],
"冒险": [
"冒险",
],
"奇幻": ["奇幻"],
"悬疑": ["悬念", "悬疑"],
"战争": ["战争"],
"灾难": ["灾难"],
"运动": ["运动"],
"动作": ["动作"],
"科幻": ["科幻"],
"黑色电影": ["黑色电影"],
"喜剧": [
"喜剧",
"Comedy",
"喜劇 Comedy",
],
"脱口秀": ["脱口秀"],
"儿童": [
"儿童",
"兒童 Kids",
],
"动画": ["动画"],
"剧情": [
"剧情",
"劇情 Drama",
],
"西部": ["西部"],
"纪录片": ["纪录片", "紀錄片 Documentary"],
"舞台艺术": ["舞台艺术", "戏曲"],
"真人秀": ["真人秀", "Reality-TV"],
"历史": ["历史"],
"犯罪": ["犯罪"],
"荒诞": ["荒诞"],
"短片": ["短片"],
"其他": [],
}
for realName, aliasList in genreMap.items():
if genreAlias in aliasList:
return realName
return "其他"
def __handle_language(self, languageAlias):
"""
语言可以按照语种进行分类，常见的语种分类包括：
1. 汉语系语言：包括汉语、粤语、闽南语、客家语等。
2. 英语系语言：包括英语、爱尔兰语、苏格兰语、威尔士语等。
3. 法语系语言：包括法语、西班牙语、葡萄牙语、意大利语等。
4. 德语系语言：包括德语、荷兰语、丹麦语、挪威语等。
5. 斯拉夫语系语言：包括俄语、乌克兰语、波兰语、捷克语等。
6. 日语系语言：包括日语、朝鲜语等。
7. 阿拉伯语系语言：包括阿拉伯语、波斯语等。
8. 印欧语系语言：包括印度语、梵语、希腊语、拉丁语等。
9. 非洲语系语言：包括斯瓦希里语、豪萨语、祖鲁语等。
10. 美洲原住民语系语言：包括纳瓦特尔语、克里语等。
"""
languageMap = {
"其他": [
"拉脱维亚语",
"斯华西里语",
"塔吉克语",
"Afrikaans",
"捷克",
"桑海语",
"Aramaic",
"无声",
"苏语",
"米斯特克语",
"silent",
"马拉雅拉姆语 Malayalam",
"阿拉伯语",
"波兰语",
"泰卢固语 Telugu",
"荷兰语",
"德清方言",
"天马电影制片厂",
"Pashtu",
"拉达克语",
"British Sign Language",
"Tonga (Tonga Islands)",
"Mongolian",
"瑞典语",
"瑞典语 Swedish",
"西班牙语",
"科萨语",
"南非荷兰语",
"意第绪语",
"孟加拉语",
"Persian",
"河北井陉话",
"泰固卢语",
"Zulu",
"奥地利",
"吉尔吉斯语",
"克罗地亚语",
"北印度语 HIndi",
"库尔德语 Kurdish",
"馬拉雅拉姆語 Malayalam",
"English",
"宗卡语",
"Silent",
"波兰语 Polish",
"Bulgarian",
"土尔其语",
"英文",
"Swahili",
"佛兰德斯语",
"班巴拉语",
"卢森堡语",
"默片",
"湘西方言",
"依地语",
"泰语",
"German",
"乌克兰语",
"阿布哈兹语",
"坎纳达语",
"罗马尼亚语 Romanian",
"马来语 Malayalam",
"法語French",
"中文",
"日本語",
"Assamese",
"冰岛语 Icelandic",
"维吾尔语",
"Serbian",
"国语",
"Telegu",
"老挝语",
"克里奥尔语",
"塞尔维亚-克罗地亚语",
"Wolof",
"少数民族语",
"世界语",
"印度语",
"孟加拉语 Bengali",
"Sicilian",
"武汉话",
"加泰罗尼亚语",
"东北方言",
"泰尔米语",
"塞尔维亚语",
"俄罗斯语 Russian",
"武汉方言",
"Bhojpuri",
"西班牙",
"菲律宾语",
"Marathi",
"泰卢固语 Telugu",
"阿尔巴尼亚语",
"西班牙語 Spanish",
"贵州独山话",
"客家话",
"土著语",
"沪语",
"瓜拉尼语",
"韃靼方言",
"门德语",
"Punjabi",
"晋语",
"泰卢固语",
"苗语",
"乌兹别克语",
"芬兰语",
"Punjab语",
"纳瓦霍语",
"印尼语",
"泰盧固語 Telugu",
"Min Nan",
"山西话",
"越剧",
"Swedish",
"Creole",
"印地语",
"赛德克语",
"台湾原住民语言",
"韩语",
"美国手语",
"客家语",
"罗马尼语",
"阿塞拜疆语",
"玛雅语",
"祖鲁语",
"瑞典语 Swedish",
"印地语 Hindi",
"陕西话",
"潮州语",
"Hindi",
"卡纳达语",
"盖尔语",
"Nahuatl",
"斯洛伐克语",
"克林贡语",
"爱沙尼亚语",
"斯洛伐克语 Slovak",
"大同方言",
"印度尼西亚语",
"塞索托语",
"Navajo",
"Ukrainian",
"Kuna",
"persian",
"乌尔都语 Ordu",
"四川话",
"Shanghainese",
"马其顿语",
"Telugu",
"南非荷兰语afrikaans",
"丹麦语",
"澳洲土著语",
"印地语 Hindi",
"维语",
"波尼族语",
"斯洛文尼亚语",
"Panjabi",
"哈萨克语",
"越南语",
"捷克语",
"高棉语",
"科西嘉语",
"Sindarin",
"Chaozhou",
"菏泽话",
"Malayalam",
"立陶宛语",
"贵州话",
"马林凯语",
"塞尔维亚－克罗地亚语",
"加利西亚语",
"Phono-Kinema",
"北印度语 Hindi",
"林加拉语",
"沃洛夫语",
"Saami",
"克伦语",
"俄语 Russia",
"印地语，乌尔杜语",
"闽南话",
"Gujarati",
"印地語 Hindi",
"塞尔维亚-克罗埃西亚语",
"Tamil",
"缅甸语",
"冰岛语",
"印地语 hindi",
"羌语",
"拉脱维亚",
"波斯语",
"阿萨姆语",
"塞尔维亚克罗地亚语",
"古吉拉特语",
"Spanish",
"Dari",
"东北话",
"俄语Russian",
"Icelandic",
"印第安苏族语",
"刚卡尼语",
"陕西方言",
"埃塞俄比亚语",
"图阿雷格语",
"北印度语",
"柏柏尔语",
"塔加拉族语",
"廈語",
"挪威",
"希腊语",
"撒丁语",
"尼泊尔语",
"粵語",
"Chinese",
"俄语",
"吉普赛语",
"法语手语",
"坦米尔语 Tamil",
"丹麦",
"Syriac",
"俄语 Russian",
"Bengali",
"闽南语",
"中国方言",
"岗德语",
"因纽特语",
"印地语 HIndi",
"俄罗斯 Russian",
"藏语",
"锡伯语",
"波斯语 Persian",
"泰米尔语 tamil",
"中文广东话",
"塔加路族语",
"希伯来语",
"古加拉特语",
"达里语",
"潮汕方言",
"无对白",
"弗拉芒语",
"塔塔尔语",
"迪尤拉语 Dyula",
"他加禄语",
"台语 Taiwanese",
"南非语",
"阿拉伯 Arabic",
"普通话",
"Washoe",
"Albanian",
"Sioux",
"印地",
"Mono",
"卡拜尔语",
"湖北方言",
"朝鲜语",
"河南话",
"挪威语",
"泰米尔 Tamil",
"阿姆哈拉语",
"亚美尼亚语",
"波斯尼亚语",
"Tamil |Hindi |Telugu",
"波斯尼亚语 Bosnian",
"西北方言",
"胶辽官话",
"粤语",
"世界语 Esperanto",
"Aidoukrou",
"Apache languages",
"蒙古语",
"Belarusian",
"厦语",
"韩国语",
"车臣语",
"匈牙利语",
"english",
"Tupi",
"上海话",
"Catalan",
"斯瓦希里语",
"大连话",
"海达语",
"马来语",
"罗马尼亚语",
"拉贾斯坦语",
"朝鲜手语",
"保加利亚语",
"俄罗斯语",
"台语",
"Tzotzil",
"Creek",
"中英字幕",
"中文 Chinese",
"乌尔都语 Urdu",
"旁遮普语",
"西西里语",
"安徽方言",
"瓦尤语",
"泰庐固语 Telugu",
"他加禄语 Tagalog",
"天津话",
"乌尔都语",
"丹麦语 Danish",
"Armenian",
"印地语 Hindi",
"菏兰语",
"客家話",
"山西方言",
"台湾手语",
"捷克语 Czech",
"巴西手语",
"夏安语",
"Tamil |Telugu |Hindi",
"蚌埠话",
"马拉提语",
"Malay",
"荷兰语 Dutch",
"温州方言",
"瑞士德语",
"僧伽罗语",
"格鲁吉亚语",
"北印度语 HINDI",
"萨米语",
"Japanese Sign Language",
"克什米尔语",
"那不勒斯语",
"傈僳语",
"hindi",
"阿拉伯语 Arabic",
"埃纳德语",
"巴斯克语",
"泰米尔语 Tamil",
"Russian",
"约鲁巴语",
"閩南語",
"英国",
"河南方言",
"Kabuverdianu",
"韩国手语",
"拉丁语",
"亚美尼加语",
"越南语 Vietnamese",
"皮钦语",
"泰卢固语Telugu",
"福建话",
"马拉雅拉姆语",
"北美印第安语",
"库尔德语",
"Hawaiian",
"Croatian",
"Esperanto",
"哈尼语",
"土耳其语",
"比印度语 Hindi",
"德语",
"意大利",
"阿尔巴尼亚语 Albanian",
"阿伊努语",
"意大利语",
"匈奴语",
"印度手语",
"马尔他语",
"泰米尔语",
"重庆方言",
"Parsee",
"Klingon",
"闽南",
"Tajik",
"四川方言",
"阿拉姆语",
"Inuktitut",
"康纳达语",
"奇楚亚语",
"台語",
"Gaelic",
"格陵兰语",
"普什图语",
"古英语",
"Athapascan languages",
"Galician",
"克丘亚语",
"毛利语",
"芬兰语 Finnish",
"泰文",
"印度语 Hindi",
"Ungwatsi",
"南京方言",
"拉丁语 Latin",
"Alsatian",
"满族语",
"满语",
"塔加路语",
"西安话",
"英語 English",
"Maya",
"坎纳达语 Kannada",
"坦纳岛西南部方言",
"天津方言",
"加里西亚语",
"索马里语",
"波利尼西亚语",
"Welsh",
],
"汉语": [
"汉语普通话",
"东北普通话",
"汉语方言",
"汉语东北话",
"云南方言",
"北京话",
"内蒙古方言",
"湖南方言",
"青岛话",
"台湾腔",
"闽南语 Min Nan",
"唐山话",
"潮州话",
"重庆话",
"山东话",
"贵州方言",
"甘肃方言",
],
"英语": [
"英語",
"英语",
"爱尔兰语",
"爱尔兰",
"爱尔兰盖尔语",
"苏格兰盖立语",
"苏格兰盖尔语",
"威尔士语",
],
"法语": [
"法语",
"葡萄牙语",
"French",
"葡萄牙语 Portuguese",
"葡萄牙语 Português",
"葡萄牙语 Portuguese",
"葡萄牙语 Portugese",
"葡萄牙",
],
"日语": [
"日语",
"日語",
],
"手语": [
"英语手语",
"法国手语",
"日本手语",
],
}
for realName, aliasList in languageMap.items():
if languageAlias in aliasList:
return realName
return "其他"
def __read_raw_data(self, ratio: float = None):
"""数据读取"""
if ratio:
movie_df = pd.read_csv(
self.movie_data_path,
index_col="MOVIE_ID",
nrows=int(sum(1 for l in open(self.movie_data_path)) * 0.1),
)
else:
movie_df = pd.read_csv(self.movie_data_path, index_col="MOVIE_ID")
person_df = pd.read_csv(self.person_data_path, index_col="PERSON_ID")
return movie_df, person_df
@timeit
def __preprocess_person(self, person_df):
"""person数据处理"""
# 保留特定列
person_df = person_df[["NAME", "SEX", "BIRTH"]]
# 根据PERSON_ID排序
person_df = person_df.sort_index()
# 获取演员的出身年份
person_df["BIRTH"] = person_df["BIRTH"].fillna("")
person_df["BIRTH_YEAR"] = person_df["BIRTH"].apply(
lambda x: x.split("-")[0] if x else None
)
person_df = person_df.drop(columns=["BIRTH"])
# 处理SEX信息
person_df["SEX"] = person_df["SEX"].apply(lambda x: 1 if x == "男" else 0)
# 填补SEX中的None
maleProbability = (
person_df["SEX"].value_counts()[1] / person_df["SEX"].value_counts().sum()
)
person_df["SEX"] = person_df["SEX"].apply(
lambda x: (
int(np.random.binomial(n=1, p=maleProbability, size=1))
if x != 0 and x != 1
else x
)
)
# 分离NAME信息
NAME = person_df["NAME"].copy()
person_df = person_df.drop(columns=["NAME"])
# 保存其他信息
personOtherInfo = {"NAME": NAME}
return person_df, personOtherInfo
@timeit
def __preprocess_movie(self, movie_df):
"""movie数据处理"""
# 保留特定列
movie_df = movie_df[
[
"NAME",
"GENRES",
"LANGUAGES",
"MINS",
"YEAR",
"DOUBAN_SCORE",
"DOUBAN_VOTES",
"ACTOR_IDS",
"DIRECTOR_IDS",
]
]
# 根据MOVIE_ID排序
movie_df = movie_df.sort_index()
# 清理无person的条目
movie_df = movie_df.dropna(subset=["ACTOR_IDS", "DIRECTOR_IDS"], how="all")
movie_df = movie_df.fillna("")
# 获取每部电影相关的person
movie_df["PERSON_IDS"] = movie_df["ACTOR_IDS"] + "|" + movie_df["DIRECTOR_IDS"]
# 只保留person的id信息
def processColumn(columnName):
# Alternative: movie_df[columnName].map(lambda x: list(filter(None,list(set(x.split("|"))))))
movie_df[columnName] = movie_df[columnName].str.split("|", expand=False)
def get_person_id(persons):
ids = []
for person in persons:
# 确保person的格式为"NAME:ID"(person数据要存在于person_df中)
if len(person.split(":")) == 2 and person.split(":")[1] != "":
ids.append(int(person.split(":")[1]))
ids = list(set(ids))
return ids
movie_df[columnName] = movie_df[columnName].map(get_person_id)
for columnName in ["ACTOR_IDS", "DIRECTOR_IDS", "PERSON_IDS"]:
processColumn(columnName)
# 去除无合作的movie NOTE: 网络至此全连通
movie_df = movie_df[movie_df["PERSON_IDS"].map(len) > 1]
# 处理DOUBAN_VOTES
movie_df["DOUBAN_VOTES"].apply(lambda x: int(x))
def remove_element(object_, val):
try:
object_.remove(val)
except ValueError:
pass
return object_
def encode(input_list, class_list):
# 统计类别数目
num_classes = len(class_list)
# 初始化编码结果
encoded = [0] * num_classes
# 遍历输入列表，将包含的类别编码为1
for x in input_list:
if x in class_list:
encoded[class_list.index(x)] = 1
return encoded
# 处理GENRES
movie_df["GENRES"] = movie_df["GENRES"].str.split("/", expand=False)
movie_df["GENRES"] = movie_df["GENRES"].apply(
lambda x: remove_element([genre.strip() for genre in x], "")
)
movie_df["GENRES"] = movie_df["GENRES"].apply(
lambda x: [self.__handle_genre(genreAlias) for genreAlias in x]
)
GENRE_LIST = list(set(movie_df["GENRES"].sum()))
movie_df["GENRES"] = movie_df["GENRES"].apply(lambda x: encode(x, GENRE_LIST))
# 处理LANGUAGES
movie_df["LANGUAGES"] = movie_df["LANGUAGES"].str.split("/", expand=False)
movie_df["LANGUAGES"] = movie_df["LANGUAGES"].apply(
lambda x: remove_element([language.strip() for language in x], "")
)
movie_df["LANGUAGES"] = movie_df["LANGUAGES"].apply(
lambda x: [self.__handle_language(languageAlias) for languageAlias in x]
)
LANGUAGE_LIST = list(set(movie_df["LANGUAGES"].sum()))
movie_df["LANGUAGES"] = movie_df["LANGUAGES"].apply(
lambda x: encode(x, LANGUAGE_LIST)
)
# 分离NAME信息
NAME = movie_df["NAME"].copy()
movie_df = movie_df.drop(columns=["NAME"])
# 保存其他信息
movieOtherInfo = {
"NAME": NAME,
"GENRE_LIST": GENRE_LIST,
"LANGUAGE_LIST": LANGUAGE_LIST,
}
return movie_df, movieOtherInfo
@timeit
def __build_graph(self, movie_df):
graph = nx.Graph()
print("根据每个电影中的相关人员，逐步添加边")
for i, (index, row) in tqdm(
enumerate(movie_df.iterrows()),
total=len(movie_df),
desc="\t 进度",
unit="%",
ncols=80,
):
person_ids = row["PERSON_IDS"]
# 遍历当前电影的每个人员ID
for i in range(len(person_ids)):
for j in range(i + 1, len(person_ids)):
# 添加这两个人员之间的边
if graph.has_edge(person_ids[i], person_ids[j]):
# 如果边已经存在，则将其权重加1
graph[person_ids[i]][person_ids[j]]["weight"] += 1
else:
# 如果边不存在，则添加一条权重为1的边
graph.add_edge(person_ids[i], person_ids[j], weight=1)
return graph
def __save_processed_data(self, data: Dict):
path = self.processed_path / Path("data.pkl")
with open(path, "wb") as f:
pickle.dump(data, f)
@timeit
def __build_featureDataFrame(
self,
movie_df: pd.DataFrame,
person_df: pd.DataFrame,
movieOtherInfo: Dict,
personOtherInfo: Dict,
):
def countProfession(row):
actorIds = row["ACTOR_IDS"]
directorIds = row["DIRECTOR_IDS"]
for directorId in directorIds:
personProfession[directorId][1] += 1
for actorId in actorIds:
personProfession[actorId][0] += 1
# 获取真实的任务ID列表
print("\t获取真实的任务ID列表")
personList = []
movie_df["PERSON_IDS"].apply(lambda ids: personList.extend(ids))
personList = list(set(personList))
# 初始化featureDataFrame
print("\t初始化featureDataFrame")
featureDataFrame = DataFrame(
None,
index=personList,
columns=[
"SEX",
"BIRTH_YEAR",
"ACTOR_COUNT",
"DIRECTOR_COUNT",
"MOVIE_COUNT",
"MEDIAN_DOUBAN_VOTES",
"MEDIAN_DOUBAN_SCORE",
"MEDIAN_MINS",
"GENRE_COUNT",
"LANGUAGE_COUNT",
"MEDIAN_YEAR",
],
)
# 初始化为列表的特征值
featureDataFrame["GENRE_COUNT"] = Series(
data=[
len(movieOtherInfo["GENRE_LIST"]) * [0]
for i in range(featureDataFrame.shape[0])
],
index=personList,
)
featureDataFrame["LANGUAGE_COUNT"] = Series(
data=[
len(movieOtherInfo["LANGUAGE_LIST"]) * [0]
for i in range(featureDataFrame.shape[0])
],
index=personList,
)
# 统计person的职业(NOTE: 相当于处理边的类型)
print("\t统计person的职业")
personProfession = {
personId: [0, 0] for personId in personList
} # NOTE: 索引0是演员次数，索引1是导演次数
movie_df.apply(countProfession, axis=1)
# 提取person_df的数据
print("\t提取person_df的数据")
for i, (index, row) in tqdm(
enumerate(person_df.iterrows()),
desc="\t 进度",
total=len(person_df),
unit="%",
ncols=80,
):
if index not in personList:
pass
else:
def get_year(string):
if not string:
return None
try:
date = datetime.datetime.strptime(string, "%Y-%m-%d")
year = date.year
except ValueError:
try:
date = datetime.datetime.strptime(string, "%Y-%m")
year = date.year
except ValueError:
try:
year = int(string)
except ValueError:
year = None
return year
sex = row["SEX"]
birthYear = get_year(row["BIRTH_YEAR"])
featureDataFrame.loc[index, "SEX"] = sex
featureDataFrame.loc[index, "BIRTH_YEAR"] = birthYear
# 获取person与movie的反向关系
print("\t获取person与movie的反向关系")
personInMovies = {personId: [] for personId in personList}
for index, row in movie_df.iterrows():
for personId in row["PERSON_IDS"]:
personInMovies[personId].append(index)
# 提取movie_df的数据
print("\t提取movie_df的数据")
featureDataFrame["MOVIE_COUNT"] = [
len(x) for x in personInMovies.values()
] # 获取参与数量
for i, (index, row) in tqdm(
enumerate(featureDataFrame.iterrows()),
total=len(featureDataFrame),
desc="\t 进度",
unit="%",
ncols=80,
):
movieIndexList = personInMovies[index]
# NOTE: 使用中位数是为了更方便的处理缺失值
featureDataFrame.loc[index, "MEDIAN_DOUBAN_VOTES"] = movie_df.loc[
movieIndexList, "DOUBAN_VOTES"
].median()
featureDataFrame.loc[index, "MEDIAN_DOUBAN_SCORE"] = movie_df.loc[
movieIndexList, "DOUBAN_SCORE"
].median()
featureDataFrame.loc[index, "MEDIAN_MINS"] = movie_df.loc[
movieIndexList, "MINS"
].median()
genreCount = [
int(x)
for x in np.array(
list(movie_df.loc[movieIndexList, "GENRES"].values)
).sum(axis=0)
]
for ind in range(len(movieOtherInfo["GENRE_LIST"])):
featureDataFrame.loc[index, "GENRE_COUNT"][ind] = genreCount[ind]
languageCount = [
int(x)
for x in np.array(
[
np.array(item)
for item in movie_df.loc[movieIndexList, "LANGUAGES"].values
]
).sum(axis=0)
]
for ind in range(len(movieOtherInfo["LANGUAGE_LIST"])):
featureDataFrame.loc[index, "LANGUAGE_COUNT"][ind] = languageCount[ind]
featureDataFrame.loc[index, "MEDIAN_YEAR"] = movie_df.loc[
movieIndexList, "YEAR"
].median()
# 边特征嵌入
print("\t边特征嵌入")
featureDataFrame["ACTOR_COUNT"] = [x[0] for x in personProfession.values()]
featureDataFrame["DIRECTOR_COUNT"] = [x[1] for x in personProfession.values()]
# 填补MEDIAN_YEAR
print("\t填补MEDIAN_YEAR")
featureDataFrame["MEDIAN_YEAR"] = featureDataFrame["MEDIAN_YEAR"].apply(
lambda year: (
np.random.uniform(low=1950, high=2015, size=1)[0]
if year < 1900.0
else year
)
)
# 填补SEX中的None NOTE: 这个填补不是因为person_df中的缺失，而是因为person_df中压根没有这个id
print("\t填补SEX")
try:
maleProbability = (
featureDataFrame["SEX"].value_counts()[1]
/ featureDataFrame["SEX"].value_counts().sum()
)
except IndexError:
maleProbability = 0.5
featureDataFrame["SEX"] = featureDataFrame["SEX"].apply(
lambda x: (
int(np.random.binomial(n=1, p=maleProbability, size=1)[0])
if x != 0 and x != 1
else x
)
)
# 填补BIRTH_YEAR中的None
# NOTE: 根据一些研究和数据，大部分演员的重要作品出现的年纪大约在 30 到 50 岁之间，这个区间可以作为一个大致的参考。
# 而重要作品的发布年份一般出现在作品年份的中位数附近
print("\t填补BIRTH_YEAR")
for index, row in featureDataFrame.loc[
featureDataFrame["BIRTH_YEAR"].isna()
].iterrows():
if row["MEDIAN_YEAR"] > 0:
featureDataFrame.loc[index, "BIRTH_YEAR"] = int(
row["MEDIAN_YEAR"] - np.random.uniform(low=30, high=50, size=1)[0]
)
# 整理dtype
print("\t整理dtype")
featureDataFrame["BIRTH_YEAR"] = featureDataFrame["BIRTH_YEAR"].astype("int64")
featureDataFrame["MEDIAN_DOUBAN_VOTES"] = featureDataFrame[
"MEDIAN_DOUBAN_VOTES"
].astype("float64")
featureDataFrame["MEDIAN_DOUBAN_SCORE"] = featureDataFrame[
"MEDIAN_DOUBAN_SCORE"
].astype("float64")
featureDataFrame["MEDIAN_MINS"] = featureDataFrame["MEDIAN_MINS"].astype(
"float64"
)
return featureDataFrame
@timeit
def __build_featureArray(self, featureDataFrame, movieOtherInfo):
featureArray = np.zeros(
(
featureDataFrame.shape[0],
9