-
Notifications
You must be signed in to change notification settings - Fork 10
/
Copy pathdata_utils.py
374 lines (345 loc) · 29.5 KB
/
data_utils.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
"""
Author: Moustafa Alzantot ([email protected])
All rights reserved.
"""
import json
import pandas as pd
from sklearn.preprocessing import OneHotEncoder
import numpy as np
# Methods to identify the distinct set of values for each column
def valid_values_from_metdata(metadata, feat_name):
""" collect distinct set of values according to the JSON metada file """
return {
k: v for k, v in enumerate(list(range(metadata[feat_name]['maxval']+1)))}
def valid_values_from_groundtruth(gt_df, feat_name):
""" collect distinct set of values according from the input data.
WE DO this ONLY for Geographic (state-dependent columns)
"""
return {
v: k for k, v in enumerate(gt_df[feat_name].unique().tolist())}
def valid_values_from_list(l):
"""
Define distinct set of values according to a list we collected from the
codebook file.
"""
return {k: v for v, k in enumerate(l)}
def preprocess_nist_data(input_file, meta_file, subsample=False):
"""
Applies pre-processing to the input data.
returns the following
original_df: the input data in the original format.
input_df: the input data after beeing formatted.
metadata: the metadata dictionary.
subsample: this argument is True only when we debug the model to work on samller datasets.
"""
original_df = pd.read_csv(input_file).astype(np.int32)
if subsample:
original_df = original_df.sample(100000)
metadata = json.loads(open(meta_file).read())
columns_list = original_df.columns.tolist()
# Make it easy iso perform one hot encoding
# Mappings areisbased on the codebook file.
col_maps = {
'SPLIT': {0: 0, 1: 1}, # from codebook
'OWNERSHP': {0: 0, 1: 1, 2: 2}, # from codebook
'VETWWI': {0: 0, 1: 1, 2: 2}, # from codebook
'RESPONDT': {0: 0, 1: 1, 2: 2}, # from codebook
'LABFORCE': {0: 0, 1: 1, 2: 2}, # from codebook
'SLREC': {1: 0, 2: 1}, # from codebook
'SSENROLL': {0: 0, 1: 1, 2: 2}, # from codebook
'SPANNAME': {0: 0, 1: 1, 2: 2, 9: 3}, # from codebook
'SCHOOL': {0: 0, 1: 1, 2: 2, 9: 3}, # from codebook
'URBAN': {0: 0, 1: 1, 2: 2}, # from codebook
'FARM': {0: 0, 1: 1, 2: 2}, # from codebook
'SEX': {1: 0, 2: 1}, # from codebook
'METRO': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4}, # from codebook
'EMPSTAT': {0: 0, 1: 1, 2: 2, 3: 3}, # from codebook
'HISPAN': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 9: 5}, # from codebook
'CITIZEN': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5}, # from codebook
'NATIVITY': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5}, # from codebook
'WARD': valid_values_from_metdata(metadata, 'WARD'), # from specs file
# from codebook
'RACE': {1: 0, 2: 1, 3: 2, 4: 3, 5: 4, 6: 5, 7: 6, 8: 7, 9: 8},
# from codebook
'WKSWORK2': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6},
'MARST': {1: 0, 2: 1, 3: 2, 4: 3, 5: 4, 6: 5}, # from codebook
'NCHLT5': {i: i for i in range(10)}, # from codebook
'GQ': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6}, # from specs file
'VETPER': {i: i for i in range(9)}, # from codebook
'HISPRULE': {i: i for i in range(9)}, # from codebook
'VET1940': {0: 0, 1: 1, 2: 2, 8: 3}, # from codebook
'UCLASSWK': {i: i for i in range(9)}, # from codebook
'HRSWORK2': {i: i for i in range(9)}, # from codebook
'SAMESEA5': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 9: 5}, # from codebook
'SAMEPLAC': {0: 0, 1: 1, 2: 2, 9: 3}, # from codebook
'GQTYPE': {i: i for i in range(10)}, # from codebook
'VETCHILD': {0: 0, 1: 1, 2: 2, 8: 3, 9: 4}, # from codebook
'CLASSWKR': {0: 0, 1: 1, 2: 2, 9: 3}, # from codebook
'VETSTAT': {0: 0, 1: 1, 2: 2, 9: 3}, # from codebook
'INCNONWG': {0: 0, 1: 1, 2: 2, 9: 3}, # from codebook
'MARRNO': {i: i for i in range(10)}, # from codebook
'MIGTYPE5': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 9: 5}, # from codebook
# from codebook
'MIGRATE5': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 8: 5, 9: 6},
# from codebook
'OWNERSHPD': {0: 0, 10: 1, 11: 2, 12: 3, 13: 4, 20: 5, 21: 6, 22: 7},
# from specs file
'FAMSIZE': valid_values_from_metdata(metadata, 'FAMSIZE'),
'SIZEPL': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6, 7: 7, 8: 8, 9: 9,
10: 10, 20: 11, 30: 12, 40: 13, 50: 14, 60: 15, 70: 16, 80: 17, 90: 18}, # from codebook
'EMPSTATD': {0: 0, 10: 1, 11: 2, 12: 3, 13: 4, 14: 5, 15: 6, 20: 7,
21: 8, 22: 9, 30: 10, 31: 11, 32: 12, 33: 13, 34: 14}, # from codebook
# values are in range[1, 52]
'WKSWORK1': valid_values_from_metdata(metadata, 'WKSWORK1'),
# SEA is one of the geographic values, we are allowd to inspect their set of values from ground truth.
# SEA is missing for codebook. Coloardo values are from https://usa.ipums.org/usa/volii/seacodes.shtml
'SEA': valid_values_from_groundtruth(original_df, 'SEA'),
# from specs file
'OCCSCORE': valid_values_from_metdata(metadata, 'OCCSCORE'),
# from metadata
'AGEMARR': valid_values_from_metdata(metadata, 'AGEMARR'),
'MIGRATE5D': {0: 0, 10: 1, 20: 2, 21: 3, 22: 4, 23: 5, 24: 6, 25: 7, 30: 8, 31: 9,
32: 10, 33: 11, 40: 12, 80: 13, 90: 14}, # from codebook
# from spec file
'MTONGUE': valid_values_from_metdata(metadata, 'MTONGUE'),
'SEI': valid_values_from_metdata(metadata, 'SEI'), # from spec file
'CLASSWKRD': {0: 0, 10: 1, 11: 2, 12: 3, 13: 4, 14: 5, 20: 6, 21: 7,
22: 8, 23: 9, 24: 10, 25: 11, 26: 12, 27: 13, 28: 14, 29: 15,
98: 16, 99: 17}, # from codebook
# from spec file
'HRSWORK1': valid_values_from_metdata(metadata, 'HRSWORK1'),
'HIGRADE': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6, 7: 7, 8: 8, 9: 9, 10: 10, 11: 11,
12: 12, 13: 13, 14: 14, 15: 15, 16: 16, 17: 17, 18: 18, 19: 19, 20: 20, 21: 21,
22: 22, 23: 23, 99: 24}, # from codebook file
'EDUC': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6, 7: 7,
8: 8, 9: 9, 10: 10, 11: 11, 99: 12}, # from codebook
# from codebook
'VETSTATD': {0: 0, 10: 1, 11: 2, 12: 3, 13: 4, 20: 5, 21: 6, 22: 7, 23: 8, 99: 9},
'CHBORN': valid_values_from_metdata(metadata, 'CHBORN'), # from specs
'GQFUNDS': {0: 0, 11: 1, 12: 2, 13: 3, 14: 4, 15: 5, 16: 6, 21: 7, 22: 8, 23: 9,
24: 10, 25: 11, 99: 12}, # from codebook
'AGEMONTH': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6, 7: 7, 8: 8, 9: 9, 10: 10,
11: 11, 12: 12, 98: 13, 99: 14}, # from codebook
'AGE': valid_values_from_metdata(metadata, 'AGE'), # max val is 108
'HISPAND': valid_values_from_list([0, 100, 102, 103, 104, 105, 106, 107,
200, 300, 401, 402, 411, 412, 413, 414, 415, 416, 417, 420, 421, 422, 423, 424,
425, 426, 427, 428, 429, 430, 431, 450, 451, 452, 453, 454, 455, 456, 457, 458,
459, 460, 465, 470, 480, 490, 491, 492, 493, 494, 495, 496, 498, 499, 900]), # from codebook
'RACED': valid_values_from_list( # from codebook
[100, 110, 120, 130, 140, 150, 200, 210, 300, 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, 328, 329, 330, 350, 351, 352,
353, 354, 355, 356, 357, 358, 359, 360, 361, 362,
370, 371, 372, 373, 374, 375, 379, 398, 399, 400,
410, 420, 500, 600, 610, 620, 630, 631, 632, 634,
640, 641, 642, 643, 650, 651, 652, 653, 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, 698, 699, 700, 801, 802, 810, 811, 812, 813,
814, 815, 816, 817, 818, 819, 820, 821, 822, 823,
824, 825, 826, 827, 830, 831, 832, 833, 834, 835,
836, 837, 838, 840, 841, 842, 845, 850, 851, 852,
853, 854, 855, 856, 860, 861, 862, 863, 864, 865,
866, 867, 868, 869, 880, 881, 882, 883, 884, 885,
886, 887, 890, 891, 892, 893, 899, 901, 902, 903,
904, 905, 906, 907, 910, 911, 912, 913, 914, 915,
916, 917, 920, 921, 922, 923, 925, 930, 931, 932,
933, 934, 935, 940, 941, 942, 943, 944, 949, 950,
951, 952, 953, 954, 955, 960, 961, 962, 963, 964,
970, 971, 972, 973, 974, 975, 976, 980, 981, 982,
983, 984, 985, 986, 989, 990, 991, 996]),
# TODO(malzantot): check
# geographic
'SUPDIST': valid_values_from_metdata(metadata, 'SUPDIST'),
# coloardo values
'METAREA': valid_values_from_groundtruth(original_df, 'METAREA'),
# from specs file
'PRESGL': valid_values_from_metdata(metadata, 'PRESGL'),
# from codebook
'MBPL': valid_values_from_list([0, 1, 2, 4, 5, 6, 8, 9, 10, 11, 12, 13, 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, 44, 45, 46, 47, 48, 49, 50, 51, 53, 54, 55, 56, 90, 99, 100, 105, 110, 115, 120, 150, 155, 160, 199, 200, 210, 250, 260, 299, 300, 400, 401, 402, 403, 404, 405, 410, 411, 412, 413, 414, 419, 420, 421, 422, 423, 424, 425, 426, 429, 430, 431, 432, 433, 434, 435, 436, 437, 438, 439, 440, 450, 451, 452, 453, 454, 455, 456, 457, 458, 459, 460, 461, 462, 463, 465, 499, 500, 501, 502, 509, 510, 511, 512, 513, 514, 515, 516, 517, 518, 519, 520, 521, 522, 523, 524, 530, 531, 532, 533, 534, 535, 536, 537, 538, 539, 540, 541, 542, 543, 544, 545, 546, 547, 548, 549, 550, 599, 600, 700, 710, 900, 950, 997, 999]),
# from codebook
'FBPL': valid_values_from_list([0, 1, 2, 4, 5, 6, 8, 9, 10, 11, 12, 13, 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, 44, 45, 46, 47, 48, 49, 50, 51, 53, 54, 55, 56, 90, 99, 100, 105, 110, 115, 120, 150, 155, 160, 199, 200, 210, 250, 260, 299, 300, 400, 401, 402, 403, 404, 405, 410, 411, 412, 413, 414, 419, 420, 421, 422, 423, 424, 425, 426, 429, 430, 431, 432, 433, 434, 435, 436, 437, 438, 439, 440, 450, 451, 452, 453, 454, 455, 456, 457, 458, 459, 460, 461, 462, 463, 465, 499, 500, 501, 502, 509, 510, 511, 512, 513, 514, 515, 516, 517, 518, 519, 520, 521, 522, 523, 524, 530, 531, 532, 533, 534, 535, 536, 537, 538, 539, 540, 541, 542, 543, 544, 545, 546, 547, 548, 549, 550, 599, 600, 700, 710, 900, 950, 997, 999]),
# from codebook
'BPL': valid_values_from_list([0, 1, 2, 4, 5, 6, 8, 9, 10, 11, 12, 13, 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, 44, 45, 46, 47, 48, 49, 50, 51, 53, 54, 55, 56, 90, 99, 100, 105, 110, 115, 120, 150, 155, 160, 199, 200, 210, 250, 260, 299, 300, 400, 401, 402, 403, 404, 405, 410, 411, 412, 413, 414, 419, 420, 421, 422, 423, 424, 425, 426, 429, 430, 431, 432, 433, 434, 435, 436, 437, 438, 439, 440, 450, 451, 452, 453, 454, 455, 456, 457, 458, 459, 460, 461, 462, 463, 465, 499, 500, 501, 502, 509, 510, 511, 512, 513, 514, 515, 516, 517, 518, 519, 520, 521, 522, 523, 524, 530, 531, 532, 533, 534, 535, 536, 537, 538, 539, 540, 541, 542, 543, 544, 545, 546, 547, 548, 549, 550, 599, 600, 700, 710, 900, 950, 997, 999]),
# from codebook
'IND1950': valid_values_from_list([0, 105, 116, 126, 206, 216, 226, 236, 239, 246, 306, 307, 308, 309, 316, 317, 318, 319, 326, 336, 337, 338, 346, 347, 348, 356, 357, 358, 367, 376, 377, 378, 379, 386, 387, 388, 399, 406, 407, 408, 409, 416, 417, 418, 419, 426, 429, 436, 437, 438, 439, 446, 448, 449, 456, 457, 458, 459, 466, 467, 468, 469, 476, 477, 478, 487, 488, 489, 499, 506, 516, 526, 527, 536, 546, 556, 567, 568, 578, 579, 586, 587, 588, 596, 597, 598, 606, 607, 608, 609, 616, 617, 618, 619, 626, 627, 636, 637, 646, 647, 656, 657, 658, 659, 667, 668, 669, 679, 686, 687, 688, 689, 696, 697, 698, 699, 716, 726, 736, 746, 756, 806, 807, 808, 816, 817, 826, 836, 846, 847, 848, 849, 856, 857, 858, 859, 868, 869, 879, 888, 896, 897, 898, 899, 906, 916, 926, 936, 946, 976, 979, 980, 982, 983, 984, 986, 987, 991, 995, 997, 998, 999]),
# State enconomic area, the list of all SEA codes are from https://usa.ipums.org/usa/volii/seacodes.shtml
'MIGSEA5': valid_values_from_list(list(range(503)) + list(range(990, 998))),
# Occ codes in 1940 cnesus from https://usa.ipums.org/usa/volii/occ1940.shtml
'OCC': valid_values_from_list([
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, 98, 99, 100, 102,
104, 106, 108, 110, 112, 114, 116, 118, 120, 122, 124, 126, 128,
130, 132, 134, 136, 156, 200, 210, 220, 222, 224, 226, 231, 236, 240,
242, 244, 246, 248, 250, 252, 254, 256, 258, 266, 270, 272, 274, 276, 278,
280, 282, 284, 286, 298, 300, 302, 304, 306, 308, 310, 312, 314, 316,
318, 320, 322, 324, 326, 327, 328, 330, 332, 334, 336, 338, 340, 342,
344, 346, 348, 350, 352, 354, 356, 358, 260, 360, 362, 364, 366, 368, 370, 372, 374,
376, 378, 380, 382, 384, 386, 388, 390, 392, 394, 396, 398, 400, 402,
404, 406, 408, 410, 412, 414, 416, 418, 420, 430, 432, 434, 436, 438,
440, 442, 444, 446, 448, 450, 452, 454, 456, 458, 460, 462, 464, 466, 468, 470,
472, 474, 476, 478, 480, 482, 484, 486, 488, 496, 500, 510, 520, 600, 602, 604, 606, 608,
610, 612, 614, 700, 710, 712, 714, 720, 730, 732, 740, 750, 760, 770, 780,
790, 792, 794, 796, 798, 844, 866, 888, 902, 904, 906, 908, 910, 988, 900, 995, 996,
998, 999
]),
# from codebook
'UOCC95': valid_values_from_list([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 17, 18, 19, 23, 24, 25, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36, 41, 42, 43, 44, 45, 46, 47, 48, 49, 51, 52, 53, 54, 55, 56, 57, 58, 59, 61, 62, 63, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 81, 82, 83, 84, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 123, 200, 201, 203, 204, 205, 210, 230, 240, 250, 260, 270, 280, 290, 300, 301, 302, 304, 305, 310, 320, 321, 322, 325, 335, 340, 341, 342, 350, 360, 365, 370, 380, 390, 400, 410, 420, 430, 450, 460, 470, 480, 490, 500, 501, 502, 503, 504, 505, 510, 511, 512, 513, 514, 515, 520, 521, 522, 523, 524, 525, 530, 531, 532, 533, 534, 535, 540, 541, 542, 543, 544, 545, 550, 551, 552, 553, 554, 555, 560, 561, 562, 563, 564, 565, 570, 571, 572, 573, 574, 575, 580, 581, 582, 583, 584, 585, 590, 591, 592, 593, 594, 595, 600, 601, 602, 603, 604, 605, 610, 611, 612, 613, 614, 615, 620, 621, 622, 623, 624, 625, 630, 631, 632, 633, 634, 635, 640, 641, 642, 643, 644, 645, 650, 660, 661, 662, 670, 671, 672, 673, 674, 675, 680, 681, 682, 683, 684, 685, 690, 700, 710, 720, 730, 731, 732, 740, 750, 751, 752, 753, 754, 760, 761, 762, 763, 764, 770, 771, 772, 773, 780, 781, 782, 783, 784, 785, 790, 810, 820, 830, 840, 910, 920, 930, 940, 950, 960, 970, 975, 980, 982, 983, 984, 986, 987, 995, 997, 999]),
# from codebook
'UIND': valid_values_from_list(list(range(132))+[995, 996, 997, 998, 999]),
# from codebook, assuming same set as UIND
'IND': valid_values_from_list(list(range(132))+[995, 996, 997, 998, 999]),
# from spec file. probably use an int ??
'DURUNEMP': valid_values_from_metdata(metadata, 'DURUNEMP'),
# from codebook
'GQTYPED': valid_values_from_list([0, 10, 20, 100, 200, 210, 211, 212, 213, 220, 221, 230, 240, 250, 260, 300, 400, 410, 411, 412, 413, 420, 421, 430, 431, 432, 440, 441, 450, 451, 452, 460, 461, 470, 471, 472, 480, 481, 482, 491, 492, 493, 494, 495, 496, 500, 501, 502, 600, 601, 602, 603, 604, 700, 701, 800, 801, 802, 803, 804, 900, 901, 910, 911, 912, 913, 914, 920, 921, 922, 923, 924, 931, 932, 933, 934, 935, 936, 937, 940, 941, 942, 943, 944, 945, 946, 947, 948, 950, 955, 960, 999]),
# from codebook
'MIGPLAC5': valid_values_from_list([0, 1, 2, 4, 5, 6, 8, 9, 10, 11, 12, 13, 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, 44, 45, 46, 47, 48, 49, 50, 51, 53, 54, 55, 56, 61, 62, 63, 64, 65, 66, 67, 68, 99, 100, 105, 110, 115, 119, 120, 150, 151, 152, 155, 160, 199, 200, 211, 212, 213, 214, 215, 216, 217, 218, 219, 250, 260, 261, 262, 263, 264, 266, 267, 305, 310, 315, 320, 325, 330, 345, 350, 360, 365, 370, 390, 400, 401, 402, 404, 405, 410, 411, 412, 413, 414, 415, 420, 421, 422, 423, 424, 425, 426, 430, 431, 432, 433, 434, 435, 436, 437, 438, 439, 440, 450, 451, 452, 453, 454, 455, 456, 457, 460, 461, 462, 465, 496, 498, 499, 500, 501, 502, 510, 511, 512, 513, 514, 515, 516, 517, 518, 520, 521, 525, 522, 523, 524, 530, 531, 532, 534, 535, 536, 537, 538, 539, 540, 541, 542, 543, 544, 548, 599, 600, 610, 612, 670, 690, 694, 699, 700, 701, 702, 710, 715, 800, 900, 911, 912, 990, 999]),
# from codebook
'HIGRADED': valid_values_from_list([0, 10, 11, 12, 20, 21, 22, 30, 31, 32, 40, 41, 42, 50, 51, 52, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90, 91, 92, 100, 101, 102, 110, 111, 112, 120, 121, 122, 130, 131, 132, 140, 141, 142, 150, 151, 152, 160, 161, 162, 170, 171, 172, 180, 181, 182, 190, 191, 192, 200, 201, 202, 210, 211, 212, 220, 221, 222, 230, 999]),
# from codebook
'EDUCD': valid_values_from_list([0, 1, 2, 10, 11, 12, 13, 14, 15, 16, 17, 20, 21, 22, 23, 24, 25, 26, 30, 40, 50, 60, 61, 62, 63, 64, 65, 70, 71, 80, 81, 82, 83, 90, 100, 101, 110, 111, 112, 113, 114, 115, 116, 999]),
# from codebook
'OCC1950': valid_values_from_list([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 17, 18, 19, 23, 24, 25, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36, 41, 42, 43, 44, 45, 46, 47, 48, 49, 51, 52, 53, 54, 55, 56, 57, 58, 59, 61, 62, 63, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 81, 82, 83, 84, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 123, 200, 201, 203, 204, 205, 210, 230, 240, 250, 260, 270, 280, 290, 300, 301, 302, 304, 305, 310, 320, 321, 322, 325, 335, 340, 341, 342, 350, 360, 365, 370, 380, 390, 400, 410, 420, 430, 450, 460, 470, 480, 490, 500, 501, 502, 503, 504, 505, 510, 511, 512, 513, 514, 515, 520, 521, 522, 523, 524, 525, 530, 531, 532, 533, 534, 535, 540, 541, 542, 543, 544, 545, 550, 551, 552, 553, 554, 555, 560, 561, 562, 563, 564, 565, 570, 571, 572, 573, 574, 575, 580, 581, 582, 583, 584, 585, 590, 591, 592, 593, 594, 595, 600, 601, 602, 603, 604, 605, 610, 611, 612, 613, 614, 615, 620, 621, 622, 623, 624, 625, 630, 631, 632, 633, 634, 635, 640, 641, 642, 643, 644, 645, 650, 660, 661, 662, 670, 671, 672, 673, 674, 675, 680, 681, 682, 683, 684, 685, 690, 700, 710, 720, 730, 731, 732, 740, 750, 751, 752, 753, 754, 760, 761, 762, 763, 764, 770, 771, 772, 773, 780, 781, 782, 783, 784, 785, 790, 810, 820, 830, 840, 910, 920, 930, 940, 950, 960, 970, 979, 980, 981, 982, 983, 984, 985, 986, 987, 990, 991, 995, 997, 999]),
# from codebook
'UOCC': valid_values_from_list([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, 98, 99, 100, 102, 104, 106, 108, 110, 112, 114, 116, 118, 120, 122, 124, 126, 128, 130, 132, 134, 136, 156, 200, 210, 220, 222, 224, 226, 236, 240, 242, 244, 246, 248, 250, 252, 254, 256, 258, 266, 270, 272, 274, 276, 278, 280, 282, 284, 286, 298, 300, 302, 304, 306, 308, 310, 312, 314, 316, 318, 320, 322, 324, 326, 327, 328, 330, 332, 334, 336, 338, 340, 342, 344, 346, 348, 350, 352, 354, 356, 358, 260, 362, 364, 366, 368, 370, 372, 374, 376, 378, 380, 382, 384, 386, 388, 390, 392, 394, 396, 398, 400, 402, 404, 406, 408, 410, 412, 414, 416, 418, 420, 430, 432, 434, 436, 438, 440, 442, 444, 446, 448, 450, 452, 454, 456, 458, 460, 462, 464, 466, 468, 470, 472, 474, 476, 478, 480, 482, 484, 486, 488, 496, 500, 510, 520, 600, 602, 604, 606, 608, 610, 612, 614, 700, 710, 712, 714, 720, 730, 732, 740, 750, 760, 770, 780, 790, 792, 794, 796, 798, 844, 866, 888, 900, 902, 904, 906, 908, 910, 988, 995, 996, 997, 998, 999]),
# COUNTY is a geographic variable where we are allowed to pick the set of values from ground truth.
# Coloardo counties from https://usa.ipums.org/usa/volii/ICPSR.shtml#colorado
'COUNTY': valid_values_from_groundtruth(original_df, 'COUNTY'),
'ENUMDIST': 'int', #
'CITYPOP': 'int', # TODO(malzantot): fix
'URBPOP': 'int', # TODO(malzantot): fix
# METAREAD is a geographic variable where we are allowed to pick the set of values from ground truth.
'METAREAD': valid_values_from_groundtruth(original_df, 'METAREAD'),
# from codebook
'MTONGUED': valid_values_from_list([0, 100, 110, 120, 130, 140, 150, 160, 200, 210, 220, 230, 240, 300, 310, 320, 400, 410, 420, 430, 440, 450, 460, 470, 500, 600, 700, 800, 810, 900, 1000, 1010, 1020, 1030, 1100, 1110, 1120, 1130, 1140, 1150, 1200, 1210, 1220, 1230, 1240, 1250, 1300, 1400, 1500, 1510, 1520, 1530, 1540, 1550, 1560, 1570, 1580, 1590, 1600, 1700, 1800, 1810, 1811, 1820, 1900, 1910, 1920, 1930, 2000, 2010, 2020, 2100, 2110, 2200, 2300, 2310, 2320, 2330, 2331, 2332, 2400, 2500, 2510, 2600, 2610, 2620, 2621, 2630, 2700, 2800, 2900, 2910, 3000, 3010, 3020, 3030, 3040, 3050, 3100, 3101, 3102, 3103, 3110, 3111, 3112, 3113, 3114, 3115, 3116, 3117, 3118, 3119, 3120, 3121, 3122, 3123, 3130, 3140, 3150, 3190, 3200, 3210, 3300, 3400, 3401, 3402, 3500, 3510, 3511, 3520, 3521, 3530, 3600, 3700, 3701, 3702, 3703, 3704, 3705, 3706, 3707, 3708, 3709, 3710, 3711, 3800, 3810, 3900, 4000, 4001, 4002, 4003, 4004, 4005, 4010, 4011, 4100, 4110, 4200, 4300, 4301, 4302, 4303, 4310, 4311, 4312, 4313, 4314, 4315, 4400, 4410, 4420, 4500, 4510, 4600, 4700, 4710, 4720, 4800, 4900, 5000, 5100, 5110, 5120, 5130, 5140, 5150, 5200, 5210, 5220, 5230, 5240, 5250, 5260, 5270, 5280, 5290, 5300, 5310, 5320, 5330, 5340, 5400, 5410, 5420, 5430, 5440, 5450, 5460, 5470, 5480, 5500, 5501, 5502, 5503, 5504, 5505, 5506, 5507, 5508, 5509, 5510, 5511, 5512, 5513, 5514, 5520, 5521, 5522, 5523, 5524, 5525, 5526, 5527, 5528, 5529, 5530, 5590, 5600, 5700, 5710, 5720, 5730, 5740, 5750, 5800, 5810, 5820, 5900, 6000, 6100, 6110, 6120, 6130, 6300, 6301, 6302, 6303, 6304, 6305, 6306, 6307, 6308, 6309, 6310, 6311, 6312, 6313, 6314, 6320, 6321, 6322, 6390, 6400, 7000, 7100, 7110, 7120, 7130, 7140, 7150, 7160, 7200, 7201, 7202, 7203, 7204, 7205, 7206, 7207, 7208, 7209, 7210, 7211, 7212, 7213, 7214, 7215, 7216, 7217, 7218, 7219, 7300, 7301, 7302, 7303, 7304, 7305, 7306, 7307, 7308, 7309, 7310, 7311, 7312, 7313, 7314, 7400, 7401, 7402, 7403, 7404, 7405, 7406, 7407, 7408, 7409, 7410, 7411, 7412, 7413, 7420, 7421, 7422, 7423, 7424, 7430, 7440, 7450, 7490, 7500, 7600, 7610, 7620, 7630, 7700, 7701, 7702, 7703, 7704, 7705, 7706, 7707, 7708, 7709, 7710, 7711, 7712, 7713, 7714, 7715, 7800, 7900, 7910, 7920, 7930, 7940, 7950, 7960, 7970, 7980, 7990, 8000, 8010, 8020, 8030, 8040, 8050, 8060, 8100, 8101, 8102, 8103, 8104, 8105, 8106, 8107, 8108, 8109, 8110, 8111, 8120, 8200, 8210, 8220, 8230, 8240, 8250, 8260, 8300, 8400, 8410, 8420, 8430, 8440, 8450, 8460, 8470, 8480, 8500, 8510, 8520, 8530, 8600, 8601, 8602, 8603, 8604, 8605, 8606, 8607, 8608, 8609, 8610, 8620, 8630, 8631, 8632, 8633, 8640, 8700, 8800, 8810, 8820, 8900, 8910, 9000, 9010, 9020, 9030, 9040, 9050, 9100, 9101, 9110, 9111, 9112, 9120, 9130, 9131, 9140, 9150, 9160, 9170, 9171, 9200, 9210, 9211, 9212, 9213, 9214, 9215, 9220, 9230, 9231, 9240, 9241, 9242, 9250, 9260, 9270, 9271, 9280, 9281, 9282, 9290, 9291, 9292, 9300, 9400, 9410, 9420, 9500, 9600, 9601, 9602, 9700, 9900, 9999]),
'EDSCOR50': 'int_v',
'MIGCITY5': 'int_v',
'NPBOSS50': 'int_v',
# CITY is a geographic variable where we are allowed to pick the set of values from ground truth.
'CITY': valid_values_from_groundtruth(original_df, 'CITY'),
'RENT': 'int_v', # int or 9999
'MIGCOUNTY': 'int_v',
'MIGMET5': 'int_v',
'ERSCOR50': 'int_v',
'MBPLD': 'void', # This one will be equal MBPL * 100
'FBPLD': 'void', # This one will be equal to FBPL * 100
'BPLD': 'void', # this one will be equal to BPL * 100
'INCWAGE': 'int_v',
'VALUEH': 'int_v'
}
#select_cols = ['SPLIT', 'OWNERSHP', 'VETWWI','VALUEH', 'CITYPOP']
#original_df = original_df[select_cols]
#col_maps = {k:col_maps[k] for k in select_cols}
output_df_columns = []
for k in columns_list:
assert k in columns_list, 'Cannot find pre-prepossing of column {}'.format(k)
v = col_maps[k]
if isinstance(v, dict):
print('Processing [ {} ] '.format(k))
mapped_col = original_df[k].map(v)
if (len(v) == 2):
output_df_columns.append(mapped_col)
else:
mapped_vals = mapped_col.values.reshape((-1, 1))
ohe_col = pd.DataFrame(
data=OneHotEncoder(n_values=len(v), sparse=False).fit_transform(mapped_vals).astype(np.float32), index=original_df.index)
output_df_columns.append(ohe_col)
elif v == 'void':
pass # skip that column
elif v == 'int':
output_df_columns.append(original_df[k] / 100.0)
elif v == 'int_v':
val_column = original_df[k].values.astype(np.float32)
is_valid_column = (
original_df[k] != metadata[k]['maxval']).astype(np.float32)
print(k, ' - ', is_valid_column.sum(),
' / ', is_valid_column.shape[0])
val_column_processed = (is_valid_column) * val_column / 100.0
output_df_columns.append(
is_valid_column.rename('{}_valid'.format(k)))
output_df_columns.append(val_column_processed)
else:
raise Exception('Invalid mapping')
output_df = pd.concat(output_df_columns, 1)
print(output_df.shape)
return original_df, output_df, metadata, col_maps, columns_list
def postprocess_data(input_data, metadata, col_maps, columns_list, greedy=True):
""" Applies post-processing to the generator model outputs """
output_df_columns = {}
cur_idx = 0
if isinstance(input_data, np.ndarray):
input_data = pd.DataFrame(data=input_data)
for k in columns_list:
assert k in col_maps, 'Coloumn mapping not found'
v = col_maps[k]
print('Post processing {}'.format(k))
if isinstance(v, dict):
col_start = cur_idx
if len(v) == 2:
# binary column
col_end = col_start + 1
output_col = (
input_data.iloc[:, col_start] > 0.5).astype(np.int32)
cur_idx += 1
else:
if greedy:
col_end = col_start + len(v)
col_ohe = input_data.iloc[:, col_start: col_end]
output_col = pd.Series(
data=np.argmax(
col_ohe.values, axis=1).astype(np.int32),
index=input_data.index)
cur_idx += len(v)
else:
col_end = col_start + 1
output_col = input_data.iloc[:, col_start].astype(np.int32)
cur_idx += 1
inv_map = {cv: ck for ck, cv in v.items()}
output_col = output_col.map(inv_map)
output_df_columns[k] = output_col.astype(np.int32)
elif v == 'int':
val_col = (100.0 * input_data.iloc[:, cur_idx]).astype(np.int32)
val_col = np.clip(val_col, 0, metadata[k]['maxval'])
output_df_columns[k] = val_col
cur_idx += 1
elif v == 'int_v':
is_valid_column = (input_data.iloc[:, cur_idx] > 0.5)
value_column = (
100.0 * input_data.iloc[:, cur_idx+1]).astype(np.int32)
value_column = np.clip(value_column, 0, metadata[k]['maxval'])
val_column_processed = (
is_valid_column * value_column) + (1-is_valid_column) * metadata[k]['maxval']
output_df_columns[k] = val_column_processed.astype(np.int32)
cur_idx += 2
elif v == 'void':
if k == 'MBPLD':
output_df_columns[k] = (
output_df_columns['MBPL'].values.reshape((-1,)) * 100).astype(np.int32)
elif k == 'FBPLD':
output_df_columns[k] = (
output_df_columns['FBPL'].values.reshape(-1,) * 100).astype(np.int32)
elif k == 'BPLD':
output_df_columns[k] = (
output_df_columns['BPL'].values.reshape((-1,)) * 100).astype(np.int32)
else:
raise Exception('Invalid mapping for column {}'.format(k))
else:
raise Exception("Invalid mapping for column {}". format(k))
output_df = pd.DataFrame(output_df_columns)
return output_df
if __name__ == '__main__':
""" Only for testing """
original_df, output_data, metadata, col_maps, columns_list = preprocess_nist_data(
'../data/colorado.csv',
'../data/colorado-specs.json', subsample=True)
output_df = postprocess_data(output_data, metadata, col_maps, columns_list, greedy=True)
assert(output_df.shape == original_df.shape)
match_count = 0
for i in range(output_df.shape[0]):
if pd.DataFrame.equals(output_df.iloc[i, :], original_df.iloc[i, :]):
match_count += 1
print('Match ratio : {}/ {}'.format(match_count, output_df.shape[0]))