-
Notifications
You must be signed in to change notification settings - Fork 398
/
Copy pathtest_target_encoder.py
507 lines (458 loc) · 19.5 KB
/
test_target_encoder.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
"""Tests for the TargetEncoder class."""
from unittest import TestCase # or `from unittest import ...` if on Python 3.4+
import category_encoders as encoders
import numpy as np
import pandas as pd
from category_encoders.datasets import load_compass, load_postcodes
import tests.helpers as th
class TestTargetEncoder(TestCase):
"""Unit tests for the Target Encoder."""
def setUp(self):
"""Set up the test case."""
self.hierarchical_cat_example = pd.DataFrame(
{
'Compass': [
'N',
'N',
'NE',
'NE',
'NE',
'SE',
'SE',
'S',
'S',
'S',
'S',
'W',
'W',
'W',
'W',
'W',
],
'Speed': [
'slow',
'slow',
'slow',
'slow',
'medium',
'medium',
'medium',
'fast',
'fast',
'fast',
'fast',
'fast',
'fast',
'fast',
'fast',
'fast',
],
'Animal': [
'Cat',
'Cat',
'Cat',
'Cat',
'Cat',
'Dog',
'Dog',
'Dog',
'Dog',
'Dog',
'Dog',
'Tiger',
'Tiger',
'Wolf',
'Wolf',
'Cougar',
],
'Plant': [
'Rose',
'Rose',
'Rose',
'Rose',
'Daisy',
'Daisy',
'Daisy',
'Daisy',
'Daffodil',
'Daffodil',
'Daffodil',
'Daffodil',
'Bluebell',
'Bluebell',
'Bluebell',
'Bluebell',
],
'target': [1, 0, 1, 1, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1],
},
columns=['Compass', 'Speed', 'Animal', 'Plant', 'target'],
)
self.hierarchical_map = {
'Compass': {'N': ('N', 'NE'), 'S': ('S', 'SE'), 'W': 'W'},
'Animal': {'Feline': ('Cat', 'Tiger', 'Cougar'), 'Canine': ('Dog', 'Wolf')},
'Plant': {
'Flower': ('Rose', 'Daisy', 'Daffodil', 'Bluebell'),
'Tree': ('Ash', 'Birch'),
},
}
def test_target_encoder(self):
"""Test that no error occurs when calling the target encoder."""
np_X = th.create_array(n_rows=100)
np_X_t = th.create_array(n_rows=50, extras=True)
np_y = np.random.randn(np_X.shape[0]) > 0.5
np_y_t = np.random.randn(np_X_t.shape[0]) > 0.5
X = th.create_dataset(n_rows=100)
X_t = th.create_dataset(n_rows=50, extras=True)
y = pd.DataFrame(np_y)
y_t = pd.DataFrame(np_y_t)
enc = encoders.TargetEncoder(verbose=1, smoothing=2, min_samples_leaf=2)
enc.fit(X, y)
th.verify_numeric(enc.transform(X_t))
th.verify_numeric(enc.transform(X_t, y_t))
def test_fit(self):
"""Test the fit method and correct values are fitted."""
k = 2
f = 10
binary_cat_example = pd.DataFrame(
{
'Trend': ['UP', 'UP', 'DOWN', 'FLAT', 'DOWN', 'UP', 'DOWN', 'FLAT', 'FLAT', 'FLAT'],
'target': [1, 1, 0, 0, 1, 0, 0, 0, 1, 1],
}
)
encoder = encoders.TargetEncoder(cols=['Trend'], min_samples_leaf=k, smoothing=f)
encoder.fit(binary_cat_example, binary_cat_example['target'])
trend_mapping = encoder.mapping['Trend']
ordinal_mapping = encoder.ordinal_encoder.category_mapping[0]['mapping']
self.assertAlmostEqual(0.4125, trend_mapping[ordinal_mapping.loc['DOWN']], delta=1e-4)
self.assertEqual(0.5, trend_mapping[ordinal_mapping.loc['FLAT']])
self.assertAlmostEqual(0.5874, trend_mapping[ordinal_mapping.loc['UP']], delta=1e-4)
def test_fit_transform(self):
"""Test the good case without unknowns or NaN values."""
k = 2
f = 10
training_data = ['UP', 'UP', 'DOWN', 'FLAT', 'DOWN', 'UP', 'DOWN', 'FLAT', 'FLAT', 'FLAT']
cases = {"with list input": training_data,
"with pd categorical input": pd.Categorical(training_data,
categories=['UP', 'FLAT', 'DOWN'])}
target = [1, 1, 0, 0, 1, 0, 0, 0, 1, 1]
for case, input_data in cases.items():
with self.subTest(case):
binary_cat_example = pd.DataFrame(
{
'Trend': input_data,
'target': target
}
)
encoder = encoders.TargetEncoder(cols=['Trend'], min_samples_leaf=k, smoothing=f)
result = encoder.fit_transform(binary_cat_example, binary_cat_example['target'])
values = result['Trend'].array
self.assertAlmostEqual(0.5874, values[0], delta=1e-4)
self.assertAlmostEqual(0.5874, values[1], delta=1e-4)
self.assertAlmostEqual(0.4125, values[2], delta=1e-4)
self.assertEqual(0.5, values[3])
def test_fit_transform_with_nan(self):
"""Test that the encoder works with NaN values."""
k = 2
f = 10
binary_cat_example = pd.DataFrame(
{
'Trend': pd.Series(
[np.nan, np.nan, 'DOWN', 'FLAT', 'DOWN', np.nan, 'DOWN', 'FLAT', 'FLAT', 'FLAT']
),
'target': [1, 1, 0, 0, 1, 0, 0, 0, 1, 1],
}
)
encoder = encoders.TargetEncoder(cols=['Trend'], min_samples_leaf=k, smoothing=f)
result = encoder.fit_transform(binary_cat_example, binary_cat_example['target'])
values = result['Trend'].array
self.assertAlmostEqual(0.5874, values[0], delta=1e-4)
self.assertAlmostEqual(0.5874, values[1], delta=1e-4)
self.assertAlmostEqual(0.4125, values[2], delta=1e-4)
self.assertEqual(0.5, values[3])
def test_handle_missing_value(self):
"""Should set the global mean value for missing values if handle_missing=value."""
df = pd.DataFrame(
{'color': ['a', 'a', 'a', 'b', 'b', 'b'], 'outcome': [1.6, 0, 0, 1, 0, 1]}
)
train = df.drop('outcome', axis=1)
target = df.drop('color', axis=1)
test = pd.Series([np.nan, 'b'], name='color')
test_target = pd.Series([0, 0])
enc = encoders.TargetEncoder(cols=['color'], handle_missing='value')
enc.fit(train, target['outcome'])
obtained = enc.transform(test, test_target)
self.assertEqual(0.6, list(obtained['color'])[0])
def test_handle_unknown_value(self):
"""Test that encoder sets the global mean value for unknown values."""
train = pd.Series(['a', 'a', 'a', 'b', 'b', 'b'], name='color')
target = pd.Series([1.6, 0, 0, 1, 0, 1], name='target')
test = pd.Series(['c', 'b'], name='color')
test_target = pd.Series([0, 0])
enc = encoders.TargetEncoder(cols=['color'], handle_unknown='value')
enc.fit(train, target)
obtained = enc.transform(test, test_target)
self.assertEqual(0.6, list(obtained['color'])[0])
def test_hierarchical_smoothing(self):
"""Test that encoder works with a hierarchical mapping."""
enc = encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=self.hierarchical_map,
cols=['Compass'],
)
result = enc.fit_transform(
self.hierarchical_cat_example, self.hierarchical_cat_example['target']
)
values = result['Compass'].array
self.assertAlmostEqual(0.6226, values[0], delta=1e-4)
self.assertAlmostEqual(0.9038, values[2], delta=1e-4)
self.assertAlmostEqual(0.1766, values[5], delta=1e-4)
self.assertAlmostEqual(0.4605, values[7], delta=1e-4)
self.assertAlmostEqual(0.4033, values[11], delta=1e-4)
def test_hierarchical_smoothing_multi(self):
"""Test that the encoder works with multiple columns."""
enc = encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=self.hierarchical_map,
cols=['Compass', 'Speed', 'Animal'],
)
result = enc.fit_transform(
self.hierarchical_cat_example, self.hierarchical_cat_example['target']
)
values = result['Compass'].array
self.assertAlmostEqual(0.6226, values[0], delta=1e-4)
self.assertAlmostEqual(0.9038, values[2], delta=1e-4)
self.assertAlmostEqual(0.1766, values[5], delta=1e-4)
self.assertAlmostEqual(0.4605, values[7], delta=1e-4)
self.assertAlmostEqual(0.4033, values[11], delta=1e-4)
values = result['Speed'].array
self.assertAlmostEqual(0.6827, values[0], delta=1e-4)
self.assertAlmostEqual(0.3962, values[4], delta=1e-4)
self.assertAlmostEqual(0.4460, values[7], delta=1e-4)
values = result['Animal'].array
self.assertAlmostEqual(0.7887, values[0], delta=1e-4)
self.assertAlmostEqual(0.3248, values[5], delta=1e-4)
self.assertAlmostEqual(0.6190, values[11], delta=1e-4)
self.assertAlmostEqual(0.1309, values[13], delta=1e-4)
self.assertAlmostEqual(0.8370, values[15], delta=1e-4)
def test_hierarchical_part_named_cols(self):
"""Test that the encoder works with a partial hierarchy."""
enc = encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=self.hierarchical_map,
cols=['Compass'],
)
result = enc.fit_transform(
self.hierarchical_cat_example, self.hierarchical_cat_example['target']
)
values = result['Compass'].array
self.assertAlmostEqual(0.6226, values[0], delta=1e-4)
self.assertAlmostEqual(0.9038, values[2], delta=1e-4)
self.assertAlmostEqual(0.1766, values[5], delta=1e-4)
self.assertAlmostEqual(0.4605, values[7], delta=1e-4)
self.assertAlmostEqual(0.4033, values[11], delta=1e-4)
values = result['Speed'].array
self.assertEqual('slow', values[0])
def test_hierarchy_pandas_index(self):
"""Test that the encoder works with a pandas index."""
df = pd.DataFrame(
{
'hello': ['a', 'b', 'c', 'a', 'a', 'b', 'c', 'd', 'd'],
'world': [0, 1, 0, 0, 1, 0, 0, 1, 1],
},
columns=pd.Index(['hello', 'world']),
)
cols = df.select_dtypes(include='object').columns
self.hierarchical_map = {
'hello': {'A': ('a', 'b'), 'B': ('c', 'd')},
}
enc = encoders.TargetEncoder(
verbose=1, smoothing=2, min_samples_leaf=2, hierarchy=self.hierarchical_map, cols=cols
)
result = enc.fit_transform(df, df['world'])
values = result['hello'].array
self.assertAlmostEqual(0.3616, values[0], delta=1e-4)
self.assertAlmostEqual(0.4541, values[1], delta=1e-4)
self.assertAlmostEqual(0.2425, values[2], delta=1e-4)
self.assertAlmostEqual(0.7425, values[7], delta=1e-4)
def test_hierarchy_single_mapping(self):
"""Test that mapping a single column works."""
enc = encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=self.hierarchical_map,
cols=['Plant'],
)
result = enc.fit_transform(
self.hierarchical_cat_example, self.hierarchical_cat_example['target']
)
values = result['Plant'].array
self.assertAlmostEqual(0.6828, values[0], delta=1e-4)
self.assertAlmostEqual(0.5, values[4], delta=1e-4)
self.assertAlmostEqual(0.5, values[8], delta=1e-4)
self.assertAlmostEqual(0.3172, values[12], delta=1e-4)
def test_hierarchy_no_mapping(self):
"""Test that a trivial hierarchy mapping label to itself works."""
hierarchical_map = {
'Plant': {
'Rose': 'Rose',
'Daisy': 'Daisy',
'Daffodil': 'Daffodil',
'Bluebell': 'Bluebell',
}
}
enc = encoders.TargetEncoder(
verbose=1, smoothing=2, min_samples_leaf=2, hierarchy=hierarchical_map, cols=['Plant']
)
result = enc.fit_transform(
self.hierarchical_cat_example, self.hierarchical_cat_example['target']
)
values = result['Plant'].array
self.assertAlmostEqual(0.6828, values[0], delta=1e-4)
self.assertAlmostEqual(0.5, values[4], delta=1e-4)
self.assertAlmostEqual(0.5, values[8], delta=1e-4)
self.assertAlmostEqual(0.3172, values[12], delta=1e-4)
def test_hierarchy_error(self):
"""Test that an error is raised when the hierarchy dictionary is invalid."""
hierarchical_map = {'Plant': {'Flower': {'Rose': ('Pink', 'Yellow', 'Red')}, 'Tree': 'Ash'}}
with self.assertRaises(ValueError):
encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=hierarchical_map,
cols=['Plant'],
)
def test_trivial_hierarchy(self):
"""Test that a trivial hierarchy works."""
trivial_hierarchical_map = {'Plant': {'Plant': ('Rose', 'Daisy', 'Daffodil', 'Bluebell')}}
enc_hier = encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=trivial_hierarchical_map,
cols=['Plant'],
)
result_hier = enc_hier.fit_transform(
self.hierarchical_cat_example, self.hierarchical_cat_example['target']
)
enc_no_hier = encoders.TargetEncoder(
verbose=1, smoothing=2, min_samples_leaf=2, cols=['Plant']
)
result_no_hier = enc_no_hier.fit_transform(
self.hierarchical_cat_example, self.hierarchical_cat_example['target']
)
pd.testing.assert_series_equal(result_hier['Plant'], result_no_hier['Plant'])
def test_hierarchy_multi_level(self):
"""Test that hierarchy works with a multi-level hierarchy."""
hierarchy_multi_level_df = pd.DataFrame(
{
'Animal': [
'Cat',
'Cat',
'Dog',
'Dog',
'Dog',
'Osprey',
'Kite',
'Kite',
'Carp',
'Carp',
'Carp',
'Clownfish',
'Clownfish',
'Lizard',
'Snake',
'Snake',
],
'target': [1, 0, 1, 1, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1],
},
columns=['Animal', 'target'],
)
hierarchy_multi_level = {
'Animal': {
'Warm-Blooded': {
'Mammals': ('Cat', 'Dog'),
'Birds': ('Osprey', 'Kite'),
'Fish': ('Carp', 'Clownfish'),
},
'Cold-Blooded': {'Reptiles': ('Lizard'), 'Amphibians': ('Snake', 'Frog')},
}
}
enc = encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=hierarchy_multi_level,
cols=['Animal'],
)
result = enc.fit_transform(hierarchy_multi_level_df, hierarchy_multi_level_df['target'])
values = result['Animal'].array
self.assertAlmostEqual(0.6261, values[0], delta=1e-4)
self.assertAlmostEqual(0.9065, values[2], delta=1e-4)
self.assertAlmostEqual(0.2556, values[5], delta=1e-4)
self.assertAlmostEqual(0.3680, values[8], delta=1e-4)
self.assertAlmostEqual(0.4626, values[11], delta=1e-4)
self.assertAlmostEqual(0.1535, values[13], delta=1e-4)
self.assertAlmostEqual(0.4741, values[14], delta=1e-4)
def test_hierarchy_columnwise_compass(self):
"""Test that hierarchy works with a columnwise hierarchy."""
X, y = load_compass()
cols = X.columns[~X.columns.str.startswith('HIER')]
HIER_cols = X.columns[X.columns.str.startswith('HIER')]
enc = encoders.TargetEncoder(
verbose=1, smoothing=2, min_samples_leaf=2, hierarchy=X[HIER_cols], cols=['compass']
)
result = enc.fit_transform(X[cols], y)
values = result['compass'].array
self.assertAlmostEqual(0.6226, values[0], delta=1e-4)
self.assertAlmostEqual(0.9038, values[2], delta=1e-4)
self.assertAlmostEqual(0.1766, values[5], delta=1e-4)
self.assertAlmostEqual(0.4605, values[7], delta=1e-4)
self.assertAlmostEqual(0.4033, values[11], delta=1e-4)
def test_hierarchy_columnwise_postcodes(self):
"""Test that hierarchy works with a columnwise hierarchy."""
X, y = load_postcodes('binary')
cols = X.columns[~X.columns.str.startswith('HIER')]
HIER_cols = X.columns[X.columns.str.startswith('HIER')]
enc = encoders.TargetEncoder(
verbose=1, smoothing=2, min_samples_leaf=2, hierarchy=X[HIER_cols], cols=['postcode']
)
result = enc.fit_transform(X[cols], y)
values = result['postcode'].array
self.assertAlmostEqual(0.8448, values[0], delta=1e-4)
def test_hierarchy_columnwise_missing_level(self):
"""Test that an error is raised when a hierarchy is given but a sub-column is missing."""
X, y = load_postcodes('binary')
HIER_cols = ['HIER_postcode_1', 'HIER_postcode_2', 'HIER_postcode_4']
with self.assertRaises(ValueError):
encoders.TargetEncoder(
verbose=1,
smoothing=2,
min_samples_leaf=2,
hierarchy=X[HIER_cols],
cols=['postcode'],
)
def test_hierarchy_mapping_no_cols(self):
"""Test that an error is raised when the hierarchy is given but no columns to encode."""
hierarchical_map = {'Compass': {'N': ('N', 'NE'), 'S': ('S', 'SE'), 'W': 'W'}}
with self.assertRaises(ValueError):
encoders.TargetEncoder(
verbose=1, smoothing=2, min_samples_leaf=2, hierarchy=hierarchical_map
)
def test_hierarchy_mapping_cols_missing(self):
"""Test that an error is raised when the dataframe is missing the hierarchy column."""
X = ['N', 'N', 'NE', 'NE', 'NE', 'SE', 'SE', 'S', 'S', 'S', 'S', 'W', 'W', 'W', 'W', 'W']
hierarchical_map = {'Compass': {'N': ('N', 'NE'), 'S': ('S', 'SE'), 'W': 'W'}}
y = [1, 0, 1, 1, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1]
enc = encoders.TargetEncoder(
verbose=1, smoothing=2, min_samples_leaf=2, hierarchy=hierarchical_map, cols=['Compass']
)
with self.assertRaises(ValueError):
enc.fit_transform(X, y)