-
Notifications
You must be signed in to change notification settings - Fork 97
/
base.py
463 lines (354 loc) · 17.5 KB
/
base.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
import matplotlib.pyplot as plt
import numpy as np
from sklearn.utils.validation import check_consistent_length
from sklearn.utils import check_matplotlib_support
from ..utils import check_is_binary
from ..metrics import (
uplift_curve, perfect_uplift_curve, uplift_auc_score,
qini_curve, perfect_qini_curve, qini_auc_score,
treatment_balance_curve, uplift_by_percentile
)
def plot_uplift_preds(trmnt_preds, ctrl_preds, log=False, bins=100):
"""Plot histograms of treatment, control and uplift predictions.
Args:
trmnt_preds (1d array-like): Predictions for all observations if they are treatment.
ctrl_preds (1d array-like): Predictions for all observations if they are control.
log (bool): Logarithm of source samples. Default is False.
bins (integer or sequence): Number of histogram bins to be used. Default is 100.
If an integer is given, bins + 1 bin edges are calculated and returned.
If bins is a sequence, gives bin edges, including left edge of first bin and right edge of last bin.
In this case, bins is returned unmodified. Default is 100.
Returns:
Object that stores computed values.
"""
# TODO: Add k as parameter: vertical line on plots
check_consistent_length(trmnt_preds, ctrl_preds)
if not isinstance(bins, int) or bins <= 0:
raise ValueError(
f'Bins should be positive integer. Invalid value for bins: {bins}')
if log:
trmnt_preds = np.log(trmnt_preds + 1)
ctrl_preds = np.log(ctrl_preds + 1)
fig, axes = plt.subplots(ncols=3, nrows=1, figsize=(20, 7))
axes[0].hist(
trmnt_preds, bins=bins, alpha=0.3, color='b', label='Treated', histtype='stepfilled')
axes[0].set_ylabel('Probability hist')
axes[0].legend()
axes[0].set_title('Treatment predictions')
axes[1].hist(
ctrl_preds, bins=bins, alpha=0.5, color='y', label='Not treated', histtype='stepfilled')
axes[1].legend()
axes[1].set_title('Control predictions')
axes[2].hist(
trmnt_preds - ctrl_preds, bins=bins, alpha=0.5, color='green', label='Uplift', histtype='stepfilled')
axes[2].legend()
axes[2].set_title('Uplift predictions')
return axes
class UpliftCurveDisplay:
"""Qini and Uplift curve visualization.
Args:
x_actual, y_actual (array (shape = [>2]), array (shape = [>2])): Points on a curve
x_baseline, y_baseline (array (shape = [>2]), array (shape = [>2])): Points on a random curve
x_perfect, y_perfect (array (shape = [>2]), array (shape = [>2])): Points on a perfect curve
random (bool): Plotting a random curve
perfect (bool): Plotting a perfect curve
estimator_name (str): Name of estimator. If None, the estimator name is not shown.
"""
def __init__(self, x_actual, y_actual, x_baseline=None,
y_baseline=None, x_perfect=None, y_perfect=None,
random=None, perfect=None, estimator_name=None):
self.x_actual = x_actual
self.y_actual = y_actual
self.x_baseline = x_baseline
self.y_baseline = y_baseline
self.x_perfect = x_perfect
self.y_perfect = y_perfect
self.random = random
self.perfect = perfect
self.estimator_name = estimator_name
def plot(self, auc_score, ax=None, name=None, title=None, **kwargs):
"""Plot visualization
Args:
auc_score (float): Area under curve.§
ax (matplotlib axes): Axes object to plot on. If `None`, a new figure and axes is created. Default is None.
name (str): Name of ROC Curve for labeling. If `None`, use the name of the estimator. Default is None.
title (str): Title plot. Default is None.
Returns:
Object that stores computed values
"""
check_matplotlib_support('UpliftCurveDisplay.plot')
name = self.estimator_name if name is None else name
line_kwargs = {}
if auc_score is not None and name is not None:
line_kwargs["label"] = f"{name} ({title} = {auc_score:0.2f})"
elif auc_score is not None:
line_kwargs["label"] = f"{title} = {auc_score:0.2f}"
elif name is not None:
line_kwargs["label"] = name
line_kwargs.update(**kwargs)
if ax is None:
fig, ax = plt.subplots()
self.line_, = ax.plot(self.x_actual, self.y_actual, **line_kwargs)
if self.random:
ax.plot(self.x_baseline, self.y_baseline, label="Random")
ax.fill_between(self.x_actual, self.y_actual, self.y_baseline, alpha=0.2)
if self.perfect:
ax.plot(self.x_perfect, self.y_perfect, label="Perfect")
ax.set_xlabel('Number targeted')
ax.set_ylabel('Number of incremental outcome')
if self.random == self.perfect:
variance = False
else:
variance = True
if len(ax.lines) > 4:
ax.lines.pop(len(ax.lines) - 1)
if variance == False:
ax.lines.pop(len(ax.lines) - 1)
if "label" in line_kwargs:
ax.legend(loc=u'upper left', bbox_to_anchor=(1, 1))
self.ax_ = ax
self.figure_ = ax.figure
return self
def plot_qini_curve(y_true, uplift, treatment,
random=True, perfect=True, negative_effect=True, ax=None, name=None, **kwargs):
"""Plot Qini curves from predictions.
Args:
y_true (1d array-like): Ground truth (correct) binary labels.
uplift (1d array-like): Predicted uplift, as returned by a model.
treatment (1d array-like): Treatment labels.
random (bool): Draw a random curve. Default is True.
perfect (bool): Draw a perfect curve. Default is True.
negative_effect (bool): If True, optimum Qini Curve contains the negative effects
(negative uplift because of campaign). Otherwise, optimum Qini Curve will not
contain the negative effects. Default is True.
ax (object): The graph on which the function will be built. Default is None.
name (string): The name of the function. Default is None.
Returns:
Object that stores computed values.
Example::
from sklift.viz import plot_qini_curve
qini_disp = plot_qini_curve(
y_test, uplift_predicted, trmnt_test,
perfect=True, name='Model name'
);
qini_disp.figure_.suptitle("Qini curve");
"""
check_matplotlib_support('plot_qini_curve')
check_consistent_length(y_true, uplift, treatment)
check_is_binary(treatment)
check_is_binary(y_true)
y_true, uplift, treatment = np.array(y_true), np.array(uplift), np.array(treatment)
x_actual, y_actual = qini_curve(y_true, uplift, treatment)
if random:
x_baseline, y_baseline = x_actual, x_actual * y_actual[-1] / len(y_true)
else:
x_baseline, y_baseline = None, None
if perfect:
x_perfect, y_perfect = perfect_qini_curve(
y_true, treatment, negative_effect)
else:
x_perfect, y_perfect = None, None
viz = UpliftCurveDisplay(
x_actual=x_actual,
y_actual=y_actual,
x_baseline=x_baseline,
y_baseline=y_baseline,
x_perfect=x_perfect,
y_perfect=y_perfect,
random=random,
perfect=perfect,
estimator_name=name,
)
auc = qini_auc_score(y_true, uplift, treatment, negative_effect)
return viz.plot(auc, ax=ax, title="AUC", **kwargs)
def plot_uplift_curve(y_true, uplift, treatment,
random=True, perfect=True, ax=None, name=None, **kwargs):
"""Plot Uplift curves from predictions.
Args:
y_true (1d array-like): Ground truth (correct) binary labels.
uplift (1d array-like): Predicted uplift, as returned by a model.
treatment (1d array-like): Treatment labels.
random (bool): Draw a random curve. Default is True.
perfect (bool): Draw a perfect curve. Default is True.
ax (object): The graph on which the function will be built. Default is None.
name (string): The name of the function. Default is None.
Returns:
Object that stores computed values.
Example::
from sklift.viz import plot_uplift_curve
uplift_disp = plot_uplift_curve(
y_test, uplift_predicted, trmnt_test,
perfect=True, name='Model name'
);
uplift_disp.figure_.suptitle("Uplift curve");
"""
check_matplotlib_support('plot_uplift_curve')
check_consistent_length(y_true, uplift, treatment)
check_is_binary(treatment)
check_is_binary(y_true)
y_true, uplift, treatment = np.array(y_true), np.array(uplift), np.array(treatment)
x_actual, y_actual = uplift_curve(y_true, uplift, treatment)
if random:
x_baseline, y_baseline = x_actual, x_actual * y_actual[-1] / len(y_true)
else:
x_baseline, y_baseline = None, None
if perfect:
x_perfect, y_perfect = perfect_uplift_curve(y_true, treatment)
else:
x_perfect, y_perfect = None, None
viz = UpliftCurveDisplay(
x_actual=x_actual,
y_actual=y_actual,
x_baseline=x_baseline,
y_baseline=y_baseline,
x_perfect=x_perfect,
y_perfect=y_perfect,
random=random,
perfect=perfect,
estimator_name=name,
)
auc = uplift_auc_score(y_true, uplift, treatment)
return viz.plot(auc, ax=ax, title="AUC", **kwargs)
def plot_uplift_by_percentile(y_true, uplift, treatment, strategy='overall',
kind='line', bins=10, string_percentiles=True):
"""Plot uplift score, treatment response rate and control response rate at each percentile.
Treatment response rate ia a target mean in the treatment group.
Control response rate is a target mean in the control group.
Uplift score is a difference between treatment response rate and control response rate.
Args:
y_true (1d array-like): Correct (true) binary target values.
uplift (1d array-like): Predicted uplift, as returned by a model.
treatment (1d array-like): Treatment labels.
strategy (string, ['overall', 'by_group']): Determines the calculating strategy. Default is 'overall'.
* ``'overall'``:
The first step is taking the first k observations of all test data ordered by uplift prediction
(overall both groups - control and treatment) and conversions in treatment and control groups
calculated only on them. Then the difference between these conversions is calculated.
* ``'by_group'``:
Separately calculates conversions in top k observations in each group (control and treatment)
sorted by uplift predictions. Then the difference between these conversions is calculated.
kind (string, ['line', 'bar']): The type of plot to draw. Default is 'line'.
* ``'line'``:
Generates a line plot.
* ``'bar'``:
Generates a traditional bar-style plot.
bins (int): Determines а number of bins (and the relative percentile) in the test data. Default is 10.
string_percentiles (bool): type of xticks: float or string to plot. Default is True (string).
Returns:
Object that stores computed values.
"""
strategy_methods = ['overall', 'by_group']
kind_methods = ['line', 'bar']
check_consistent_length(y_true, uplift, treatment)
check_is_binary(treatment)
check_is_binary(y_true)
n_samples = len(y_true)
if strategy not in strategy_methods:
raise ValueError(f'Response rate supports only calculating methods in {strategy_methods},'
f' got {strategy}.')
if kind not in kind_methods:
raise ValueError(f'Function supports only types of plots in {kind_methods},'
f' got {kind}.')
if not isinstance(bins, int) or bins <= 0:
raise ValueError(
f'Bins should be positive integer. Invalid value bins: {bins}')
if bins >= n_samples:
raise ValueError(
f'Number of bins = {bins} should be smaller than the length of y_true {n_samples}')
if not isinstance(string_percentiles, bool):
raise ValueError(f'string_percentiles flag should be bool: True or False.'
f' Invalid value string_percentiles: {string_percentiles}')
df = uplift_by_percentile(y_true, uplift, treatment, strategy=strategy,
std=True, total=True, bins=bins, string_percentiles=False)
percentiles = df.index[:bins].values.astype(float)
response_rate_trmnt = df.loc[percentiles, 'response_rate_treatment'].values
std_trmnt = df.loc[percentiles, 'std_treatment'].values
response_rate_ctrl = df.loc[percentiles, 'response_rate_control'].values
std_ctrl = df.loc[percentiles, 'std_control'].values
uplift_score = df.loc[percentiles, 'uplift'].values
std_uplift = df.loc[percentiles, 'std_uplift'].values
uplift_weighted_avg = df.loc['total', 'uplift']
check_consistent_length(percentiles, response_rate_trmnt,
response_rate_ctrl, uplift_score,
std_trmnt, std_ctrl, std_uplift)
if kind == 'line':
_, axes = plt.subplots(ncols=1, nrows=1, figsize=(8, 6))
axes.errorbar(percentiles, response_rate_trmnt, yerr=std_trmnt,
linewidth=2, color='forestgreen', label='treatment\nresponse rate')
axes.errorbar(percentiles, response_rate_ctrl, yerr=std_ctrl,
linewidth=2, color='orange', label='control\nresponse rate')
axes.errorbar(percentiles, uplift_score, yerr=std_uplift,
linewidth=2, color='red', label='uplift')
axes.fill_between(percentiles, response_rate_trmnt,
response_rate_ctrl, alpha=0.1, color='red')
if np.amin(uplift_score) < 0:
axes.axhline(y=0, color='black', linewidth=1)
if string_percentiles: # string percentiles for plotting
percentiles_str = [f"0-{percentiles[0]:.0f}"] + \
[f"{percentiles[i]:.0f}-{percentiles[i + 1]:.0f}" for i in range(len(percentiles) - 1)]
axes.set_xticks(percentiles)
axes.set_xticklabels(percentiles_str, rotation=45)
else:
axes.set_xticks(percentiles)
axes.legend(loc='upper right')
axes.set_title(
f'Uplift by percentile\nweighted average uplift = {uplift_weighted_avg:.4f}')
axes.set_xlabel('Percentile')
axes.set_ylabel(
'Uplift = treatment response rate - control response rate')
else: # kind == 'bar'
delta = percentiles[0]
fig, axes = plt.subplots(ncols=1, nrows=2, figsize=(8, 6), sharex=True, sharey=True)
fig.text(0.04, 0.5, 'Uplift = treatment response rate - control response rate',
va='center', ha='center', rotation='vertical')
axes[1].bar(np.array(percentiles) - delta / 6, response_rate_trmnt, delta / 3,
yerr=std_trmnt, color='forestgreen', label='treatment\nresponse rate')
axes[1].bar(np.array(percentiles) + delta / 6, response_rate_ctrl, delta / 3,
yerr=std_ctrl, color='orange', label='control\nresponse rate')
axes[0].bar(np.array(percentiles), uplift_score, delta / 1.5,
yerr=std_uplift, color='red', label='uplift')
axes[0].legend(loc='upper right')
axes[0].tick_params(axis='x', bottom=False)
axes[0].axhline(y=0, color='black', linewidth=1)
axes[0].set_title(
f'Uplift by percentile\nweighted average uplift = {uplift_weighted_avg:.4f}')
if string_percentiles: # string percentiles for plotting
percentiles_str = [f"0-{percentiles[0]:.0f}"] + \
[f"{percentiles[i]:.0f}-{percentiles[i + 1]:.0f}" for i in range(len(percentiles) - 1)]
axes[1].set_xticks(percentiles)
axes[1].set_xticklabels(percentiles_str, rotation=45)
else:
axes[1].set_xticks(percentiles)
axes[1].legend(loc='upper right')
axes[1].axhline(y=0, color='black', linewidth=1)
axes[1].set_xlabel('Percentile')
axes[1].set_title('Response rate by percentile')
return axes
def plot_treatment_balance_curve(uplift, treatment, random=True, winsize=0.1):
"""Plot Treatment Balance curve.
Args:
uplift (1d array-like): Predicted uplift, as returned by a model.
treatment (1d array-like): Treatment labels.
random (bool): Draw a random curve. Default is True.
winsize (float): Size of the sliding window to apply. Should be between 0 and 1, extremes excluded. Default is 0.1.
Returns:
Object that stores computed values.
"""
check_consistent_length(uplift, treatment)
check_is_binary(treatment)
if (winsize <= 0) or (winsize >= 1):
raise ValueError(
'winsize should be between 0 and 1, extremes excluded')
x_tb, y_tb = treatment_balance_curve(
uplift, treatment, winsize=int(len(uplift) * winsize))
_, ax = plt.subplots(ncols=1, nrows=1, figsize=(14, 7))
ax.plot(x_tb, y_tb, label='Model', color='b')
if random:
y_tb_random = np.average(treatment) * np.ones_like(x_tb)
ax.plot(x_tb, y_tb_random, label='Random', color='black')
ax.fill_between(x_tb, y_tb, y_tb_random, alpha=0.2, color='b')
ax.legend()
ax.set_title('Treatment balance curve')
ax.set_xlabel('Percentage targeted')
ax.set_ylabel('Balance: treatment / (treatment + control)')
return ax