-
Notifications
You must be signed in to change notification settings - Fork 4.2k
/
pandas_doctests_test.py
865 lines (828 loc) · 37.7 KB
/
pandas_doctests_test.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
#
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import sys
import unittest
import pandas as pd
from apache_beam.dataframe import doctests
from apache_beam.dataframe.frames import PD_VERSION
from apache_beam.dataframe.pandas_top_level_functions import _is_top_level_function
@unittest.skipIf(sys.platform == 'win32', '[BEAM-10626]')
class DoctestTest(unittest.TestCase):
def test_ndframe_tests(self):
# IO methods are tested in io_test.py
skip_writes = {
f'pandas.core.generic.NDFrame.{name}': ['*']
for name in dir(pd.core.generic.NDFrame) if name.startswith('to_')
}
result = doctests.testmod(
pd.core.generic,
use_beam=False,
report=True,
wont_implement_ok={
'pandas.core.generic.NDFrame.head': ['*'],
'pandas.core.generic.NDFrame.shift': [
'df.shift(periods=3)',
'df.shift(periods=3, fill_value=0)',
],
'pandas.core.generic.NDFrame.tail': ['*'],
'pandas.core.generic.NDFrame.take': ['*'],
'pandas.core.generic.NDFrame.values': ['*'],
'pandas.core.generic.NDFrame.tz_localize': [
"s.tz_localize('CET', ambiguous='infer')",
# np.array is not a deferred object. This use-case is possible
# with a deferred Series though, which is tested in
# frames_test.py
"s.tz_localize('CET', ambiguous=np.array([True, True, False]))",
],
'pandas.core.generic.NDFrame.truncate': [
# These inputs rely on tail (wont implement, order
# sensitive) for verification
"df.tail()",
"df.truncate(before=pd.Timestamp('2016-01-05'),\n"
" after=pd.Timestamp('2016-01-10')).tail()",
"df.truncate('2016-01-05', '2016-01-10').tail()",
"df.loc['2016-01-05':'2016-01-10', :].tail()"
],
'pandas.core.generic.NDFrame.replace': [
"s.replace([1, 2], method='bfill')",
# Relies on method='pad'
"s.replace('a')",
# Relies on method='pad'
# value=None is not valid for pandas < 1.4
"s.replace('a', None)",
# Implicitly uses method='pad', but output doesn't rely on that
# behavior. Verified indepently in
# frames_test.py::DeferredFrameTest::test_replace
"df.replace(regex={r'^ba.$': 'new', 'foo': 'xyz'})"
],
'pandas.core.generic.NDFrame.fillna': [
'df.fillna(method=\'ffill\')',
'df.fillna(method="ffill")',
'df.fillna(value=values, limit=1)',
],
'pandas.core.generic.NDFrame.sort_values': ['*'],
'pandas.core.generic.NDFrame.mask': [
'df.where(m, -df) == np.where(m, df, -df)'
],
'pandas.core.generic.NDFrame.where': [
'df.where(m, -df) == np.where(m, df, -df)'
],
'pandas.core.generic.NDFrame.interpolate': ['*'],
'pandas.core.generic.NDFrame.resample': ['*'],
'pandas.core.generic.NDFrame.rolling': ['*'],
# argsort wont implement
'pandas.core.generic.NDFrame.abs': [
'df.loc[(df.c - 43).abs().argsort()]',
],
'pandas.core.generic.NDFrame.reindex': ['*'],
'pandas.core.generic.NDFrame.pct_change': ['*'],
'pandas.core.generic.NDFrame.asof': ['*'],
'pandas.core.generic.NDFrame.infer_objects': ['*'],
'pandas.core.generic.NDFrame.ewm': ['*'],
'pandas.core.generic.NDFrame.expanding': ['*'],
'pandas.core.generic.NDFrame.get': ['*'],
},
not_implemented_ok={
'pandas.core.generic.NDFrame.asof': ['*'],
'pandas.core.generic.NDFrame.at_time': ['*'],
'pandas.core.generic.NDFrame.between_time': ['*'],
'pandas.core.generic.NDFrame.ewm': ['*'],
'pandas.core.generic.NDFrame.expanding': ['*'],
'pandas.core.generic.NDFrame.flags': ['*'],
'pandas.core.generic.NDFrame.rank': ['*'],
'pandas.core.generic.NDFrame.reindex_like': ['*'],
'pandas.core.generic.NDFrame.replace': ['*'],
'pandas.core.generic.NDFrame.sample': ['*'],
'pandas.core.generic.NDFrame.set_flags': ['*'],
'pandas.core.generic.NDFrame.squeeze': ['*'],
'pandas.core.generic.NDFrame.truncate': ['*'],
},
skip={
# Internal test
'pandas.core.generic.NDFrame._set_axis_name': ['*'],
# Fails to construct test series. asfreq is not implemented anyway.
'pandas.core.generic.NDFrame.asfreq': ['*'],
'pandas.core.generic.NDFrame.astype': ['*'],
'pandas.core.generic.NDFrame.convert_dtypes': ['*'],
'pandas.core.generic.NDFrame.copy': ['*'],
'pandas.core.generic.NDFrame.droplevel': ['*'],
'pandas.core.generic.NDFrame.get': ['*'],
'pandas.core.generic.NDFrame.rank': [
# Modified dataframe
'df'
],
'pandas.core.generic.NDFrame.rename': [
# Seems to be an upstream bug. The actual error has a different
# message:
# TypeError: Index(...) must be called with a collection of
# some kind, 2 was passed
# pandas doctests only verify the type of exception
'df.rename(2)'
],
# For pandas >= 1.4, rename is changed to _rename
'pandas.core.generic.NDFrame._rename': [
# Seems to be an upstream bug. The actual error has a different
# message:
# TypeError: Index(...) must be called with a collection of
# some kind, 2 was passed
# pandas doctests only verify the type of exception
'df.rename(2)'
],
# Tests rely on setting index
'pandas.core.generic.NDFrame.rename_axis': ['*'],
# Raises right exception, but testing framework has matching issues.
'pandas.core.generic.NDFrame.replace': [
"df.replace({'a string': 'new value', True: False}) # raises"
],
'pandas.core.generic.NDFrame.squeeze': ['*'],
# NameError
'pandas.core.generic.NDFrame.resample': ['df'],
# Skipped so we don't need to install natsort
'pandas.core.generic.NDFrame.sort_values': [
'from natsort import index_natsorted',
'df.sort_values(\n'
' by="time",\n'
' key=lambda x: np.argsort(index_natsorted(df["time"]))\n'
')'
],
**skip_writes
})
self.assertEqual(result.failed, 0)
def test_dataframe_tests(self):
result = doctests.testmod(
pd.core.frame,
use_beam=False,
report=True,
wont_implement_ok={
'pandas.core.frame.DataFrame.T': ['*'],
'pandas.core.frame.DataFrame.cummax': ['*'],
'pandas.core.frame.DataFrame.cummin': ['*'],
'pandas.core.frame.DataFrame.cumsum': ['*'],
'pandas.core.frame.DataFrame.cumprod': ['*'],
'pandas.core.frame.DataFrame.diff': ['*'],
'pandas.core.frame.DataFrame.fillna': [
'df.fillna(method=\'ffill\')',
'df.fillna(method="ffill")',
'df.fillna(value=values, limit=1)',
],
'pandas.core.frame.DataFrame.items': ['*'],
'pandas.core.frame.DataFrame.itertuples': ['*'],
'pandas.core.frame.DataFrame.iterrows': ['*'],
'pandas.core.frame.DataFrame.iteritems': ['*'],
# default keep is 'first'
'pandas.core.frame.DataFrame.nlargest': [
"df.nlargest(3, 'population')",
"df.nlargest(3, ['population', 'GDP'])",
"df.nlargest(3, 'population', keep='last')"
],
'pandas.core.frame.DataFrame.nsmallest': [
"df.nsmallest(3, 'population')",
"df.nsmallest(3, ['population', 'GDP'])",
"df.nsmallest(3, 'population', keep='last')",
],
'pandas.core.frame.DataFrame.replace': [
"s.replace([1, 2], method='bfill')",
# Relies on method='pad'
"s.replace('a')",
# Relies on method='pad'
# value=None is not valid for pandas < 1.4
"s.replace('a', None)",
# Implicitly uses method='pad', but output doesn't rely on that
# behavior. Verified indepently in
# frames_test.py::DeferredFrameTest::test_replace
"df.replace(regex={r'^ba.$': 'new', 'foo': 'xyz'})"
],
'pandas.core.frame.DataFrame.to_records': ['*'],
'pandas.core.frame.DataFrame.to_dict': ['*'],
'pandas.core.frame.DataFrame.to_numpy': ['*'],
'pandas.core.frame.DataFrame.to_string': ['*'],
'pandas.core.frame.DataFrame.transpose': ['*'],
'pandas.core.frame.DataFrame.shape': ['*'],
'pandas.core.frame.DataFrame.shift': [
'df.shift(periods=3)',
'df.shift(periods=3, fill_value=0)',
],
'pandas.core.frame.DataFrame.unstack': ['*'],
'pandas.core.frame.DataFrame.memory_usage': ['*'],
'pandas.core.frame.DataFrame.info': ['*'],
# Not equal to df.agg('mode', axis='columns', numeric_only=True)
# because there can be multiple columns if a row has more than one
# mode
'pandas.core.frame.DataFrame.mode': [
"df.mode(axis='columns', numeric_only=True)"
],
'pandas.core.frame.DataFrame.append': [
'df.append(df2, ignore_index=True)',
"for i in range(5):\n" +
" df = df.append({'A': i}, ignore_index=True)",
],
'pandas.core.frame.DataFrame.sort_index': ['*'],
'pandas.core.frame.DataFrame.sort_values': ['*'],
'pandas.core.frame.DataFrame.melt': [
"df.melt(id_vars=['A'], value_vars=['B'])",
"df.melt(id_vars=['A'], value_vars=['B', 'C'])",
"df.melt(col_level=0, id_vars=['A'], value_vars=['B'])",
"df.melt(id_vars=[('A', 'D')], value_vars=[('B', 'E')])",
"df.melt(id_vars=['A'], value_vars=['B'],\n" +
" var_name='myVarname', value_name='myValname')"
],
# Most keep= options are order-sensitive
'pandas.core.frame.DataFrame.drop_duplicates': ['*'],
'pandas.core.frame.DataFrame.duplicated': [
'df.duplicated()',
"df.duplicated(keep='last')",
"df.duplicated(subset=['brand'])",
],
'pandas.core.frame.DataFrame.reindex': ['*'],
'pandas.core.frame.DataFrame.dot': [
# reindex not supported
's2 = s.reindex([1, 0, 2, 3])',
],
'pandas.core.frame.DataFrame.resample': ['*'],
'pandas.core.frame.DataFrame.values': ['*'],
},
not_implemented_ok={
'pandas.core.frame.DataFrame.transform': [
# str arg not supported. Tested with np.sum in
# frames_test.py::DeferredFrameTest::test_groupby_transform_sum
"df.groupby('Date')['Data'].transform('sum')",
],
'pandas.core.frame.DataFrame.melt': ['*'],
'pandas.core.frame.DataFrame.reindex_axis': ['*'],
'pandas.core.frame.DataFrame.round': [
'df.round(decimals)',
],
# We should be able to support pivot and pivot_table for categorical
# columns
'pandas.core.frame.DataFrame.pivot': ['*'],
# Trivially elementwise for axis=columns. Relies on global indexing
# for axis=rows.
# Difficult to determine proxy, need to inspect function
'pandas.core.frame.DataFrame.apply': ['*'],
# Cross-join not implemented
'pandas.core.frame.DataFrame.merge': [
"df1.merge(df2, how='cross')"
],
# TODO(BEAM-11711)
'pandas.core.frame.DataFrame.set_index': [
"df.set_index([s, s**2])",
],
'pandas.core.frame.DataFrame.set_axis': [
"df.set_axis(range(0,2), axis='index')",
],
# TODO(BEAM-12495)
'pandas.core.frame.DataFrame.value_counts': [
'df.value_counts(dropna=False)'
],
},
skip={
# DataFrame construction from a dictionary and
# Series requires using the len() function, which
# is a non-deferred operation that we do not allow
'pandas.core.frame.DataFrame': [
'pd.DataFrame(data=d, index=[0, 1, 2, 3])',
],
# s2 created with reindex
'pandas.core.frame.DataFrame.dot': [
'df.dot(s2)',
],
'pandas.core.frame.DataFrame.resample': ['df'],
'pandas.core.frame.DataFrame.asfreq': ['*'],
# Throws NotImplementedError when modifying df
'pandas.core.frame.DataFrame.axes': [
# Returns deferred index.
'df.axes',
],
# Skipped because the relies on loc to set cells in df2
'pandas.core.frame.DataFrame.compare': ['*'],
'pandas.core.frame.DataFrame.cov': [
# Relies on setting entries ahead of time.
"df.loc[df.index[:5], 'a'] = np.nan",
"df.loc[df.index[5:10], 'b'] = np.nan",
'df.cov(min_periods=12)',
],
'pandas.core.frame.DataFrame.rename': [
# Returns deferred index.
'df.index',
'df.rename(index=str).index',
],
'pandas.core.frame.DataFrame.set_index': [
# TODO(BEAM-11711): This could pass in the index as
# a DeferredIndex, and we should fail it as order-sensitive.
"df.set_index([pd.Index([1, 2, 3, 4]), 'year'])",
],
'pandas.core.frame.DataFrame.set_axis': [
# This should pass as set_axis(axis='columns')
# and fail with set_axis(axis='index')
"df.set_axis(['a', 'b', 'c'], axis='index')"
],
'pandas.core.frame.DataFrame.to_markdown': ['*'],
'pandas.core.frame.DataFrame.to_parquet': ['*'],
# Raises right exception, but testing framework has matching issues.
# Tested in `frames_test.py`.
'pandas.core.frame.DataFrame.insert': [
'df',
'df.insert(1, "newcol", [99, 99])',
'df.insert(0, "col1", [100, 100], allow_duplicates=True)'
],
'pandas.core.frame.DataFrame.to_records': [
'df.index = df.index.rename("I")',
'index_dtypes = f"<S{df.index.str.len().max()}"', # 1.x
'index_dtypes = "<S{}".format(df.index.str.len().max())', #0.x
'df.to_records(index_dtypes=index_dtypes)',
],
# These tests use the static method pd.pivot_table, which doesn't
# actually raise NotImplementedError
'pandas.core.frame.DataFrame.pivot_table': ['*'],
# Expected to raise a ValueError, but we raise NotImplementedError
'pandas.core.frame.DataFrame.pivot': [
"df.pivot(index='foo', columns='bar', values='baz')"
],
'pandas.core.frame.DataFrame.append': [
'df',
# pylint: disable=line-too-long
"pd.concat([pd.DataFrame([i], columns=['A']) for i in range(5)],\n"
" ignore_index=True)"
],
'pandas.core.frame.DataFrame.eval': ['df'],
'pandas.core.frame.DataFrame.melt': [
"df.columns = [list('ABC'), list('DEF')]", "df"
],
'pandas.core.frame.DataFrame.merge': [
# Order-sensitive index, checked in frames_test.py.
"df1.merge(df2, left_on='lkey', right_on='rkey')",
"df1.merge(df2, left_on='lkey', right_on='rkey',\n"
" suffixes=('_left', '_right'))",
"df1.merge(df2, how='left', on='a')",
],
# Raises right exception, but testing framework has matching issues.
'pandas.core.frame.DataFrame.replace': [
"df.replace({'a string': 'new value', True: False}) # raises"
],
'pandas.core.frame.DataFrame.to_sparse': ['type(df)'],
# Skipped because "seen_wont_implement" is reset before getting to
# these calls, so the NameError they raise is not ignored.
'pandas.core.frame.DataFrame.T': [
'df1_transposed.dtypes', 'df2_transposed.dtypes'
],
'pandas.core.frame.DataFrame.transpose': [
'df1_transposed.dtypes', 'df2_transposed.dtypes'
],
# Skipped because the relies on iloc to set a cell to NA. Test is
# replicated in frames_test::DeferredFrameTest::test_applymap.
'pandas.core.frame.DataFrame.applymap': [
'df_copy.iloc[0, 0] = pd.NA',
"df_copy.applymap(lambda x: len(str(x)), na_action='ignore')",
],
# Skipped so we don't need to install natsort
'pandas.core.frame.DataFrame.sort_values': [
'from natsort import index_natsorted',
'df.sort_values(\n'
' by="time",\n'
' key=lambda x: np.argsort(index_natsorted(df["time"]))\n'
')'
],
# Mode that we don't yet support, documentation added in pandas
# 1.2.0 (https://github.com/pandas-dev/pandas/issues/35912)
'pandas.core.frame.DataFrame.aggregate': [
"df.agg(x=('A', max), y=('B', 'min'), z=('C', np.mean))"
],
})
self.assertEqual(result.failed, 0)
def test_series_tests(self):
result = doctests.testmod(
pd.core.series,
use_beam=False,
report=True,
wont_implement_ok={
'pandas.core.series.Series.__array__': ['*'],
'pandas.core.series.Series.array': ['*'],
'pandas.core.series.Series.cummax': ['*'],
'pandas.core.series.Series.cummin': ['*'],
'pandas.core.series.Series.cumsum': ['*'],
'pandas.core.series.Series.cumprod': ['*'],
'pandas.core.series.Series.diff': ['*'],
'pandas.core.series.Series.dot': [
's.dot(arr)', # non-deferred result
],
'pandas.core.series.Series.fillna': [
'df.fillna(method=\'ffill\')',
'df.fillna(method="ffill")',
'df.fillna(value=values, limit=1)',
],
'pandas.core.series.Series.info': ['*'],
'pandas.core.series.Series.items': ['*'],
'pandas.core.series.Series.iteritems': ['*'],
# default keep is 'first'
'pandas.core.series.Series.nlargest': [
"s.nlargest()",
"s.nlargest(3)",
"s.nlargest(3, keep='last')",
],
'pandas.core.series.Series.memory_usage': ['*'],
'pandas.core.series.Series.nsmallest': [
"s.nsmallest()",
"s.nsmallest(3)",
"s.nsmallest(3, keep='last')",
],
'pandas.core.series.Series.pop': ['*'],
'pandas.core.series.Series.searchsorted': ['*'],
'pandas.core.series.Series.shift': [
'df.shift(periods=3)',
'df.shift(periods=3, fill_value=0)',
],
'pandas.core.series.Series.take': ['*'],
'pandas.core.series.Series.to_dict': ['*'],
'pandas.core.series.Series.unique': ['*'],
'pandas.core.series.Series.unstack': ['*'],
'pandas.core.series.Series.values': ['*'],
'pandas.core.series.Series.view': ['*'],
'pandas.core.series.Series.append': [
's1.append(s2, ignore_index=True)',
],
'pandas.core.series.Series.replace': [
"s.replace([1, 2], method='bfill')",
# Relies on method='pad'
"s.replace('a')",
# Relies on method='pad'
# value=None is not valid for pandas < 1.4
"s.replace('a', None)",
# Implicitly uses method='pad', but output doesn't rely on that
# behavior. Verified indepently in
# frames_test.py::DeferredFrameTest::test_replace
"df.replace(regex={r'^ba.$': 'new', 'foo': 'xyz'})"
],
'pandas.core.series.Series.sort_index': ['*'],
'pandas.core.series.Series.sort_values': ['*'],
'pandas.core.series.Series.argmax': ['*'],
'pandas.core.series.Series.argmin': ['*'],
'pandas.core.series.Series.drop_duplicates': [
's.drop_duplicates()',
"s.drop_duplicates(keep='last')",
],
'pandas.core.series.Series.reindex': ['*'],
'pandas.core.series.Series.autocorr': ['*'],
'pandas.core.series.Series.repeat': ['s.repeat([1, 2, 3])'],
'pandas.core.series.Series.resample': ['*'],
'pandas.core.series.Series': ['ser.iloc[0] = 999'],
},
not_implemented_ok={
'pandas.core.series.Series.transform': [
# str arg not supported. Tested with np.sum in
# frames_test.py::DeferredFrameTest::test_groupby_transform_sum
"df.groupby('Date')['Data'].transform('sum')",
],
'pandas.core.series.Series.groupby': [
'ser.groupby(["a", "b", "a", "b"]).mean()',
'ser.groupby(["a", "b", "a", np.nan]).mean()',
'ser.groupby(["a", "b", "a", np.nan], dropna=False).mean()',
],
},
skip={
# Relies on setting values with iloc
'pandas.core.series.Series': ['ser', 'r'],
'pandas.core.series.Series.groupby': [
# TODO(BEAM-11393): This example requires aligning two series
# with non-unique indexes. It only works in pandas because
# pandas can recognize the indexes are identical and elide the
# alignment.
'ser.groupby(ser > 100).mean()',
],
'pandas.core.series.Series.asfreq': ['*'],
# error formatting
'pandas.core.series.Series.append': [
's1.append(s2, verify_integrity=True)',
],
'pandas.core.series.Series.cov': [
# Differs in LSB on jenkins.
"s1.cov(s2)",
],
# Skipped idxmax/idxmin due an issue with the test framework
'pandas.core.series.Series.idxmin': ['s.idxmin()'],
'pandas.core.series.Series.idxmax': ['s.idxmax()'],
'pandas.core.series.Series.duplicated': ['*'],
'pandas.core.series.Series.set_axis': ['*'],
'pandas.core.series.Series.nonzero': ['*'],
'pandas.core.series.Series.pop': ['ser'], # testing side effect
# Raises right exception, but testing framework has matching issues.
'pandas.core.series.Series.replace': [
"df.replace({'a string': 'new value', True: False}) # raises"
],
'pandas.core.series.Series.searchsorted': [
# This doctest seems to be incorrectly parsed.
"x = pd.Categorical(['apple', 'bread', 'bread',"
],
'pandas.core.series.Series.to_csv': ['*'],
'pandas.core.series.Series.to_markdown': ['*'],
'pandas.core.series.Series.update': ['*'],
'pandas.core.series.Series.view': [
# Inspection after modification.
's'
],
'pandas.core.series.Series.resample': ['df'],
})
self.assertEqual(result.failed, 0)
def test_string_tests(self):
if PD_VERSION < (1, 2):
module = pd.core.strings
else:
# Definitions were moved to accessor in pandas 1.2.0
module = pd.core.strings.accessor
module_name = module.__name__
result = doctests.testmod(
module,
use_beam=False,
wont_implement_ok={
# These methods can accept deferred series objects, but not lists
f'{module_name}.StringMethods.cat': [
"s.str.cat(['A', 'B', 'C', 'D'], sep=',')",
"s.str.cat(['A', 'B', 'C', 'D'], sep=',', na_rep='-')",
"s.str.cat(['A', 'B', 'C', 'D'], na_rep='-')"
],
f'{module_name}.StringMethods.repeat': [
's.str.repeat(repeats=[1, 2, 3])'
],
f'{module_name}.str_repeat': ['s.str.repeat(repeats=[1, 2, 3])'],
# get_dummies pandas examples are not casted to CategoricalDtype
# Must be CategoricalDtype to work in Beam
f'{module_name}.StringMethods.get_dummies': ['*'],
f'{module_name}.str_get_dummies': ['*'],
f'{module_name}.StringMethods': ['s.str.split("_")'],
f'{module_name}.StringMethods.rsplit': ['*'],
f'{module_name}.StringMethods.split': ['*'],
},
skip={
# count() on Series with a NaN produces mismatched type if we
# have a NaN-only partition.
f'{module_name}.StringMethods.count': ["s.str.count('a')"],
f'{module_name}.str_count': ["s.str.count('a')"],
# Bad test strings in pandas 1.1.x
f'{module_name}.str_replace': [
"pd.Series(['foo', 'fuz', np.nan]).str.replace('f', repr)"
],
f'{module_name}.StringMethods.replace': [
"pd.Series(['foo', 'fuz', np.nan]).str.replace('f', repr)"
],
# output has incorrect formatting in 1.2.x
f'{module_name}.StringMethods.extractall': ['*']
})
self.assertEqual(result.failed, 0)
def test_datetime_tests(self):
# TODO(BEAM-10721)
indexes_accessors_result = doctests.testmod(
pd.core.indexes.accessors,
use_beam=False,
skip={
'pandas.core.indexes.accessors.TimedeltaProperties': [
# Seems like an upstream bug. The property is 'second'
'seconds_series.dt.seconds'
],
# TODO(BEAM-12530): Test data creation fails for these
# s = pd.Series(pd.to_timedelta(np.arange(5), unit="d"))
# pylint: disable=line-too-long
'pandas.core.indexes.accessors.DatetimeProperties.to_pydatetime': [
'*'
],
'pandas.core.indexes.accessors.TimedeltaProperties.components': [
'*'
],
'pandas.core.indexes.accessors.TimedeltaProperties.to_pytimedelta': [
'*'
],
# pylint: enable=line-too-long
})
datetimelike_result = doctests.testmod(
pd.core.arrays.datetimelike, use_beam=False)
datetime_result = doctests.testmod(
pd.core.arrays.datetimes,
use_beam=False,
wont_implement_ok={
'pandas.core.arrays.datetimes.DatetimeArray.to_period': ['*'],
# All tz_localize tests use unsupported values for ambiguous=
# Verified seperately in
# frames_test.py::DeferredFrameTest::test_dt_tz_localize_*
'pandas.core.arrays.datetimes.DatetimeArray.tz_localize': ['*'],
},
not_implemented_ok={
# Verifies index version of this method
'pandas.core.arrays.datetimes.DatetimeArray.to_period': [
'df.index.to_period("M")'
],
})
self.assertEqual(indexes_accessors_result.failed, 0)
self.assertEqual(datetimelike_result.failed, 0)
self.assertEqual(datetime_result.failed, 0)
def test_indexing_tests(self):
result = doctests.testmod(
pd.core.indexing,
use_beam=False,
skip={
'pandas.core.indexing._IndexSlice': ['*'],
'pandas.core.indexing.IndexingMixin.at': ['*'],
'pandas.core.indexing.IndexingMixin.iat': ['*'],
'pandas.core.indexing.IndexingMixin.iloc': ['*'],
'pandas.core.indexing.IndexingMixin.loc': ['*'],
'pandas.core.indexing._AtIndexer': ['*'],
'pandas.core.indexing._LocIndexer': ['*'],
'pandas.core.indexing._iAtIndexer': ['*'],
'pandas.core.indexing._iLocIndexer': ['*'],
})
self.assertEqual(result.failed, 0)
def test_groupby_tests(self):
result = doctests.testmod(
pd.core.groupby.groupby,
use_beam=False,
wont_implement_ok={
'pandas.core.groupby.groupby.GroupBy.head': ['*'],
'pandas.core.groupby.groupby.GroupBy.tail': ['*'],
'pandas.core.groupby.groupby.GroupBy.nth': ['*'],
'pandas.core.groupby.groupby.GroupBy.cumcount': ['*'],
'pandas.core.groupby.groupby.GroupBy.resample': ['*'],
},
not_implemented_ok={
'pandas.core.groupby.groupby.GroupBy.ngroup': ['*'],
'pandas.core.groupby.groupby.GroupBy.sample': ['*'],
'pandas.core.groupby.groupby.GroupBy.rank': ['*'],
'pandas.core.groupby.groupby.GroupBy.nth': [
"df.groupby('A', as_index=False).nth(1)",
],
},
skip={
# Uses iloc to mutate a DataFrame
'pandas.core.groupby.groupby.GroupBy.resample': [
'df.iloc[2, 0] = 5',
'df',
],
# df is reassigned
'pandas.core.groupby.groupby.GroupBy.rank': ['df'],
# TODO: Raise wont implement for list passed as a grouping column
# Currently raises unhashable type: list
'pandas.core.groupby.groupby.GroupBy.ngroup': [
'df.groupby(["A", [1,1,2,3,2,1]]).ngroup()'
],
})
self.assertEqual(result.failed, 0)
result = doctests.testmod(
pd.core.groupby.generic,
use_beam=False,
wont_implement_ok={
# Returns an array by default, not a Series. WontImplement
# (non-deferred)
'pandas.core.groupby.generic.SeriesGroupBy.unique': ['*'],
# TODO: Is take actually deprecated?
'pandas.core.groupby.generic.DataFrameGroupBy.take': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.take': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.nsmallest': [
"s.nsmallest(3, keep='last')",
"s.nsmallest(3)",
"s.nsmallest()",
],
'pandas.core.groupby.generic.SeriesGroupBy.nlargest': [
"s.nlargest(3, keep='last')",
"s.nlargest(3)",
"s.nlargest()",
],
'pandas.core.groupby.generic.DataFrameGroupBy.diff': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.diff': ['*'],
'pandas.core.groupby.generic.DataFrameGroupBy.hist': ['*'],
'pandas.core.groupby.generic.DataFrameGroupBy.fillna': [
'df.fillna(method=\'ffill\')',
'df.fillna(method="ffill")',
'df.fillna(value=values, limit=1)',
],
'pandas.core.groupby.generic.SeriesGroupBy.fillna': [
'df.fillna(method=\'ffill\')',
'df.fillna(method="ffill")',
'df.fillna(value=values, limit=1)',
],
},
not_implemented_ok={
'pandas.core.groupby.generic.DataFrameGroupBy.idxmax': ['*'],
'pandas.core.groupby.generic.DataFrameGroupBy.idxmin': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.transform': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.idxmax': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.idxmin': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.apply': ['*'],
},
skip={
'pandas.core.groupby.generic.SeriesGroupBy.cov': [
# Floating point comparison fails
's1.cov(s2)',
],
'pandas.core.groupby.generic.DataFrameGroupBy.cov': [
# Mutates input DataFrame with loc
# TODO: Replicate in frames_test.py
"df.loc[df.index[:5], 'a'] = np.nan",
"df.loc[df.index[5:10], 'b'] = np.nan",
"df.cov(min_periods=12)",
],
# These examples rely on grouping by a list
'pandas.core.groupby.generic.SeriesGroupBy.aggregate': ['*'],
'pandas.core.groupby.generic.DataFrameGroupBy.aggregate': ['*'],
'pandas.core.groupby.generic.SeriesGroupBy.transform': [
# Dropping invalid columns during a transform is unsupported.
'grouped.transform(lambda x: (x - x.mean()) / x.std())'
],
'pandas.core.groupby.generic.DataFrameGroupBy.transform': [
# Dropping invalid columns during a transform is unsupported.
'grouped.transform(lambda x: (x - x.mean()) / x.std())'
],
# Skipped idxmax/idxmin due an issue with the test framework
'pandas.core.groupby.generic.SeriesGroupBy.idxmin': ['s.idxmin()'],
'pandas.core.groupby.generic.SeriesGroupBy.idxmax': ['s.idxmax()'],
# Uses as_index, which is currently not_implemented
'pandas.core.groupby.generic.DataFrameGroupBy.value_counts': [
"df.groupby('gender', as_index=False).value_counts()",
# pylint: disable=line-too-long
"df.groupby('gender', as_index=False).value_counts(normalize=True)",
],
})
self.assertEqual(result.failed, 0)
def test_top_level(self):
tests = {
name: func.__doc__
for (name, func) in pd.__dict__.items()
if _is_top_level_function(func) and getattr(func, '__doc__', None)
}
# IO methods are tested in io_test.py
skip_reads = {name: ['*'] for name in dir(pd) if name.startswith('read_')}
result = doctests.teststrings(
tests,
use_beam=False,
report=True,
not_implemented_ok={
'concat': ['pd.concat([s1, s2], ignore_index=True)'],
'crosstab': ['*'],
'cut': ['*'],
'eval': ['*'],
'get_dummies': ['*'],
'infer_freq': ['*'],
'lreshape': ['*'],
'melt': ['*'],
'merge': ["df1.merge(df2, how='cross')"],
'merge_asof': ['*'],
'pivot': ['*'],
'pivot_table': ['*'],
'qcut': ['*'],
'reset_option': ['*'],
'set_eng_float_format': ['*'],
'set_option': ['*'],
'to_numeric': ['*'],
'to_timedelta': ['*'],
'unique': ['*'],
'wide_to_long': ['*'],
},
wont_implement_ok={
'factorize': ['*'],
'to_datetime': ['s.head()'],
'to_pickle': ['*'],
'melt': [
"pd.melt(df, id_vars=['A'], value_vars=['B'])",
"pd.melt(df, id_vars=['A'], value_vars=['B', 'C'])",
"pd.melt(df, col_level=0, id_vars=['A'], value_vars=['B'])",
"pd.melt(df, id_vars=[('A', 'D')], value_vars=[('B', 'E')])",
"pd.melt(df, id_vars=['A'], value_vars=['B'],\n" +
" var_name='myVarname', value_name='myValname')"
],
},
skip={
# error formatting
'concat': ['pd.concat([df5, df6], verify_integrity=True)'],
# doctest DeprecationWarning
'melt': ['df'],
# Order-sensitive re-indexing.
'merge': [
"df1.merge(df2, left_on='lkey', right_on='rkey')",
"df1.merge(df2, left_on='lkey', right_on='rkey',\n"
" suffixes=('_left', '_right'))",
"df1.merge(df2, how='left', on='a')",
],
# Not an actual test.
'option_context': ['*'],
'factorize': ['codes', 'uniques'],
# Bad top-level use of un-imported function.
'merge_ordered': [
'merge_ordered(df1, df2, fill_method="ffill", left_by="group")'
],
# Expected error.
'pivot': ["df.pivot(index='foo', columns='bar', values='baz')"],
# Never written.
'to_pickle': ['os.remove("./dummy.pkl")'],
**skip_reads
})
self.assertEqual(result.failed, 0)
if __name__ == '__main__':
unittest.main()