/
_ctsf.py
578 lines (504 loc) · 24.3 KB
/
_ctsf.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
"""Composable time series forest."""
__author__ = ["mloning", "AyushmaanSeth"]
__all__ = ["ComposableTimeSeriesForestClassifier"]
import numbers
import numpy as np
from joblib import Parallel, delayed
from sklearn.ensemble._base import _partition_estimators
from sklearn.ensemble._forest import (
_generate_unsampled_indices,
_get_n_samples_bootstrap,
)
from sklearn.pipeline import Pipeline
from sklearn.tree import DecisionTreeClassifier
from sklearn.utils import compute_sample_weight
from sklearn.utils.multiclass import check_classification_targets
from sktime.base._panel.forest._composable import BaseTimeSeriesForest
from sktime.classification.base import BaseClassifier
from sktime.transformations.panel.summarize import RandomIntervalFeatureExtractor
from sktime.utils.slope_and_trend import _slope
from sktime.utils.validation.panel import check_X, check_X_y
from sktime.utils.warnings import warn
class ComposableTimeSeriesForestClassifier(BaseTimeSeriesForest, BaseClassifier):
"""Time Series Forest Classifier as described in [1]_.
A time series forest is an adaptation of the random
forest for time-series data. It that fits a number of decision tree
classifiers on various sub-samples of a transformed dataset and uses
averaging to improve the predictive accuracy and control over-fitting.
The sub-sample size is always the same as the original input sample size
but the samples are drawn with replacement if ``bootstrap=True`` (default).
Parameters
----------
estimator : Pipeline
A pipeline consisting of series-to-tabular transformations
and a decision tree classifier as final estimator.
n_estimators : integer, optional (default=200)
The number of trees in the forest.
max_depth : integer or None, optional (default=None)
The maximum depth of the tree. If None, then nodes are expanded until
all leaves are pure or until all leaves contain less than
min_samples_split samples.
min_samples_split : int, float, optional (default=2)
The minimum number of samples required to split an internal node:
- If int, then consider ``min_samples_split`` as the minimum number.
- If float, then ``min_samples_split`` is a fraction and
``ceil(min_samples_split * n_samples)`` are the minimum
number of samples for each split.
min_samples_leaf : int, float, optional (default=1)
The minimum number of samples required to be at a leaf node.
A split point at any depth will only be considered if it leaves at
least ``min_samples_leaf`` training samples in each of the left and
right branches. This may have the effect of smoothing the model,
especially in regression.
- If int, then consider ``min_samples_leaf`` as the minimum number.
- If float, then ``min_samples_leaf`` is a fraction and
``ceil(min_samples_leaf * n_samples)`` are the minimum
number of samples for each node.
min_weight_fraction_leaf : float, optional (default=0.)
The minimum weighted fraction of the sum total of weights (of all
the input samples) required to be at a leaf node. Samples have
equal weight when sample_weight is not provided.
max_features : int, float, string or None, optional (default=None)
The number of features to consider when looking for the best split:
- If int, then consider ``max_features`` features at each split.
- If float, then ``max_features`` is a fraction and
``int(max_features * n_features)`` features are considered at each
split.
- If "auto", then ``max_features=sqrt(n_features)``.
- If "sqrt", then ``max_features=sqrt(n_features)`` (same as "auto").
- If "log2", then ``max_features=log2(n_features)``.
- If None, then ``max_features=n_features``.
Note: the search for a split does not stop until at least one
valid partition of the node samples is found, even if it requires to
effectively inspect more than ``max_features`` features.
max_leaf_nodes : int or None, optional (default=None)
Grow trees with ``max_leaf_nodes`` in best-first fashion.
Best nodes are defined as relative reduction in impurity.
If None then unlimited number of leaf nodes.
min_impurity_decrease : float, optional (default=0.)
A node will be split if this split induces a decrease of the impurity
greater than or equal to this value.
The weighted impurity decrease equation is the following::
N_t / N * (impurity - N_t_R / N_t * right_impurity
- N_t_L / N_t * left_impurity)
where ``N`` is the total number of samples, ``N_t`` is the number of
samples at the current node, ``N_t_L`` is the number of samples in the
left child, and ``N_t_R`` is the number of samples in the right child.
``N``, ``N_t``, ``N_t_R`` and ``N_t_L`` all refer to the weighted sum,
if ``sample_weight`` is passed.
bootstrap : boolean, optional (default=False)
Whether bootstrap samples are used when building trees.
oob_score : bool (default=False)
Whether to use out-of-bag samples to estimate
the generalization accuracy.
n_jobs : int or None, optional (default=None)
The number of jobs to run in parallel for both ``fit`` and ``predict``.
``None`` means 1 unless in a :obj:`joblib.parallel_backend` context.
``-1`` means using all processors.
random_state : int, RandomState instance or None, optional (default=None)
If int, random_state is the seed used by the random number generator;
If RandomState instance, random_state is the random number generator;
If None, the random number generator is the RandomState instance used
by ``np.random``.
verbose : int, optional (default=0)
Controls the verbosity when fitting and predicting.
warm_start : bool, optional (default=False)
When set to ``True``, reuse the solution of the previous call to fit
and add more estimators to the ensemble, otherwise, just fit a whole
new forest.
class_weight : dict, list of dicts, "balanced", "balanced_subsample" or \
None, optional (default=None)
Weights associated with classes in the form ``{class_label: weight}``.
If not given, all classes are supposed to have weight one. For
multi-output problems, a list of dicts can be provided in the same
order as the columns of y.
Note that for multioutput (including multilabel) weights should be
defined for each class of every column in its own dict. For example,
for four-class multilabel classification weights should be
[{0: 1, 1: 1}, {0: 1, 1: 5}, {0: 1, 1: 1}, {0: 1, 1: 1}] instead of
[{1:1}, {2:5}, {3:1}, {4:1}].
The "balanced" mode uses the values of y to automatically adjust
weights inversely proportional to class frequencies in the input data
as ``n_samples / (n_classes * np.bincount(y))``
The "balanced_subsample" mode is the same as "balanced" except that
weights are computed based on the bootstrap sample for every tree
grown.
For multi-output, the weights of each column of y will be multiplied.
Note that these weights will be multiplied with sample_weight (passed
through the fit method) if sample_weight is specified.
max_samples : int or float, default=None
If bootstrap is True, the number of samples to draw from X
to train each base estimator.
- If None (default), then draw ``X.shape[0]`` samples.
- If int, then draw ``max_samples`` samples.
- If float, then draw ``max_samples * X.shape[0]`` samples. Thus,
``max_samples`` should be in the interval ``(0, 1)``.
Attributes
----------
estimators_ : list of DecisionTreeClassifier
The collection of fitted sub-estimators.
classes_ : array of shape = [n_classes] or a list of such arrays
The classes labels (single output problem), or a list of arrays of
class labels (multi-output problem).
n_classes_ : int or list
The number of classes (single output problem), or a list containing the
number of classes for each output (multi-output problem).
n_columns : int
The number of features when ``fit`` is performed.
n_outputs_ : int
The number of outputs when ``fit`` is performed.
feature_importances_ : data frame of shape = [n_timepoints, n_features]
The normalised feature values at each time index of
the time series forest
oob_score_ : float
Score of the training dataset obtained using an out-of-bag estimate.
oob_decision_function_ : array of shape = [n_samples, n_classes]
Decision function computed with out-of-bag estimate on the training
set. If n_estimators is small it might be possible that a data point
was never left out during the bootstrap. In this case,
``oob_decision_function_`` might contain NaN.
References
----------
.. [1] Deng et. al, A time series forest for classification and feature extraction,
Information Sciences, 239:2013.
Examples
--------
>>> from sktime.classification.ensemble import ComposableTimeSeriesForestClassifier
>>> from sktime.classification.kernel_based import RocketClassifier
>>> from sktime.datasets import load_unit_test
>>> X_train, y_train = load_unit_test(split="train") # doctest: +SKIP
>>> X_test, y_test = load_unit_test(split="test") # doctest: +SKIP
>>> clf = ComposableTimeSeriesForestClassifier(
... RocketClassifier(num_kernels=100),
... n_estimators=10,
... ) # doctest: +SKIP
>>> clf.fit(X_train, y_train) # doctest: +SKIP
ComposableTimeSeriesForestClassifier(...)
>>> y_pred = clf.predict(X_test) # doctest: +SKIP
"""
_tags = {
# packaging info
# --------------
"authors": ["mloning", "AyushmaanSeth"],
"maintainers": ["AyushmaanSeth"],
# estimator type
# --------------
"X_inner_mtype": "nested_univ", # nested pd.DataFrame
# capabilities
# --------------
"capability:feature_importance": True,
}
def __init__(
self,
estimator=None,
n_estimators=100,
max_depth=None,
min_samples_split=2,
min_samples_leaf=1,
min_weight_fraction_leaf=0.0,
max_features=None,
max_leaf_nodes=None,
min_impurity_decrease=0.0,
bootstrap=False,
oob_score=False,
n_jobs=None,
random_state=None,
verbose=0,
warm_start=False,
class_weight=None,
max_samples=None,
):
self.estimator = estimator
# Assign values, even though passed on to base estimator below,
# necessary here for cloning
self.max_depth = max_depth
self.min_samples_split = min_samples_split
self.min_samples_leaf = min_samples_leaf
self.min_weight_fraction_leaf = min_weight_fraction_leaf
self.max_features = max_features
self.max_leaf_nodes = max_leaf_nodes
self.min_impurity_decrease = min_impurity_decrease
self.max_samples = max_samples
# Pass on params.
super().__init__(
base_estimator=None,
n_estimators=n_estimators,
estimator_params=None,
bootstrap=bootstrap,
oob_score=oob_score,
n_jobs=n_jobs,
random_state=random_state,
verbose=verbose,
warm_start=warm_start,
class_weight=class_weight,
max_samples=max_samples,
)
BaseClassifier.__init__(self)
# We need to add is-fitted state when inheriting from scikit-learn
self._is_fitted = False
def fit(self, X, y, **kwargs):
"""Wrap fit to call BaseClassifier.fit.
This is a fix to get around the problem with multiple inheritance. The problem
is that if we just override _fit, this class inherits the fit from the sklearn
class BaseTimeSeriesForest. This is the simplest solution, albeit a little
hacky.
"""
return BaseClassifier.fit(self, X=X, y=y, **kwargs)
def predict(self, X, **kwargs) -> np.ndarray:
"""Wrap predict to call BaseClassifier.predict."""
return BaseClassifier.predict(self, X=X, **kwargs)
def predict_proba(self, X, **kwargs) -> np.ndarray:
"""Wrap predict_proba to call BaseClassifier.predict_proba."""
return BaseClassifier.predict_proba(self, X=X, **kwargs)
def _fit(self, X, y):
BaseTimeSeriesForest._fit(self, X=X, y=y)
def _validate_estimator(self):
if not isinstance(self.n_estimators, numbers.Integral):
raise ValueError(
"n_estimators must be an integer, "
"got {}.".format(type(self.n_estimators))
)
if self.n_estimators <= 0:
raise ValueError(
"n_estimators must be greater than zero, "
"got {}.".format(self.n_estimators)
)
# Set base estimator
if self.estimator is None:
# Set default time series forest
features = [np.mean, np.std, _slope]
steps = [
(
"transform",
RandomIntervalFeatureExtractor(
n_intervals="sqrt",
features=features,
random_state=self.random_state,
),
),
("clf", DecisionTreeClassifier(random_state=self.random_state)),
]
self._estimator = Pipeline(steps)
else:
# else check given estimator is a pipeline with prior
# transformations and final decision tree
if not isinstance(self.estimator, Pipeline):
raise ValueError("`estimator` must be pipeline with transforms.")
if not isinstance(self.estimator.steps[-1][1], DecisionTreeClassifier):
raise ValueError(
"Last step in `estimator` must be DecisionTreeClassifier."
)
self._estimator = self.estimator
# Set parameters according to naming in pipeline
estimator_params = {
"max_depth": self.max_depth,
"min_samples_split": self.min_samples_split,
"min_samples_leaf": self.min_samples_leaf,
"min_weight_fraction_leaf": self.min_weight_fraction_leaf,
"max_features": self.max_features,
"max_leaf_nodes": self.max_leaf_nodes,
"min_impurity_decrease": self.min_impurity_decrease,
}
final_estimator = self._estimator.steps[-1][0]
self.estimator_params = {
f"{final_estimator}__{pname}": pval
for pname, pval in estimator_params.items()
}
# Set renamed estimator parameters
for pname, pval in self.estimator_params.items():
self.__setattr__(pname, pval)
def _predict(self, X):
"""Predict class for X.
The predicted class of an input sample is a vote by the trees in
the forest, weighted by their probability estimates. That is,
the predicted class is the one with highest mean probability
estimate across the trees.
Parameters
----------
X : array-like or sparse matrix of shape (n_samples, n_features)
The input samples. Internally, its dtype will be converted to
``dtype=np.float32``. If a sparse matrix is provided, it will be
converted into a sparse ``csr_matrix``.
Returns
-------
y : array-like of shape (n_samples,) or (n_samples, n_outputs)
The predicted classes.
"""
proba = self.predict_proba(X)
if self.n_outputs_ == 1:
return self.classes_.take(np.argmax(proba, axis=1), axis=0)
else:
n_samples = proba[0].shape[0]
# all dtypes should be the same, so just take the first
class_type = self.classes_[0].dtype
predictions = np.empty((n_samples, self.n_outputs_), dtype=class_type)
for k in range(self.n_outputs_):
predictions[:, k] = self.classes_[k].take(
np.argmax(proba[k], axis=1), axis=0
)
return predictions
def predict_log_proba(self, X):
"""Predict class log-probabilities for X.
The predicted class log-probabilities of an input sample is computed as
the log of the mean predicted class probabilities of the trees in the
forest.
Parameters
----------
X : array-like or sparse matrix of shape (n_samples, n_features)
The input samples. Internally, its dtype will be converted to
``dtype=np.float32``. If a sparse matrix is provided, it will be
converted into a sparse ``csr_matrix``.
Returns
-------
p : array of shape (n_samples, n_classes), or a list of n_outputs
such arrays if n_outputs > 1.
The class probabilities of the input samples. The order of the
classes corresponds to that in the attribute ``classes_``.
"""
proba = self.predict_proba(X)
if self.n_outputs_ == 1:
return np.log(proba)
else:
for k in range(self.n_outputs_):
proba[k] = np.log(proba[k])
return proba
def _predict_proba(self, X):
"""Predict class probabilities for X.
The predicted class probabilities of an input sample are computed as
the mean predicted class probabilities of the trees in the forest. The
class probability of a single tree is the fraction of samples of the
same class in a leaf.
Parameters
----------
X : array-like or sparse matrix of shape = [n_samples, n_features]
The input samples. Internally, its dtype will be converted to
``dtype=np.float32``. If a sparse matrix is provided, it will be
converted into a sparse ``csr_matrix``.
Returns
-------
p : array of shape = [n_samples, n_classes], or a list of n_outputs
such arrays if n_outputs > 1.
The class probabilities of the input samples. The order of the
classes corresponds to that in the attribute ``classes_``.
"""
# Check data
self.check_is_fitted()
X = check_X(X, enforce_univariate=True)
X = self._validate_X_predict(X)
# Assign chunk of trees to jobs
n_jobs, _, _ = _partition_estimators(self.n_estimators, self.n_jobs)
all_proba = Parallel(n_jobs=n_jobs, verbose=self.verbose)(
delayed(e.predict_proba)(X) for e in self.estimators_
)
return np.sum(all_proba, axis=0) / len(self.estimators_)
def _set_oob_score(self, X, y):
"""Compute out-of-bag score."""
check_X_y(X, y)
check_X(X, enforce_univariate=True)
n_classes_ = self.n_classes_
n_samples = y.shape[0]
oob_decision_function = []
oob_score = 0.0
predictions = [
np.zeros((n_samples, n_classes_[k])) for k in range(self.n_outputs_)
]
n_samples_bootstrap = _get_n_samples_bootstrap(n_samples, self.max_samples)
for estimator in self.estimators_:
final_estimator = estimator.steps[-1][1]
unsampled_indices = _generate_unsampled_indices(
final_estimator.random_state, n_samples, n_samples_bootstrap
)
p_estimator = estimator.predict_proba(X.iloc[unsampled_indices, :])
if self.n_outputs_ == 1:
p_estimator = [p_estimator]
for k in range(self.n_outputs_):
predictions[k][unsampled_indices, :] += p_estimator[k]
for k in range(self.n_outputs_):
if (predictions[k].sum(axis=1) == 0).any():
warn(
"Some inputs do not have OOB scores. "
"This probably means too few trees were used "
"to compute any reliable oob estimates.",
obj=self,
)
decision = predictions[k] / predictions[k].sum(axis=1)[:, np.newaxis]
oob_decision_function.append(decision)
oob_score += np.mean(y[:, k] == np.argmax(predictions[k], axis=1), axis=0)
if self.n_outputs_ == 1:
self.oob_decision_function_ = oob_decision_function[0]
else:
self.oob_decision_function_ = oob_decision_function
self.oob_score_ = oob_score / self.n_outputs_
# TODO - Implement this abstract method properly.
def _set_oob_score_and_attributes(self, X, y):
raise NotImplementedError("Not implemented.")
def _validate_y_class_weight(self, y):
check_classification_targets(y)
y = np.copy(y)
expanded_class_weight = None
if self.class_weight is not None:
y_original = np.copy(y)
self.classes_ = []
self.n_classes_ = []
y_store_unique_indices = np.zeros(y.shape, dtype=int)
for k in range(self.n_outputs_):
classes_k, y_store_unique_indices[:, k] = np.unique(
y[:, k], return_inverse=True
)
self.classes_.append(classes_k)
self.n_classes_.append(classes_k.shape[0])
y = y_store_unique_indices
if self.class_weight is not None:
valid_presets = ("balanced", "balanced_subsample")
if isinstance(self.class_weight, str):
if self.class_weight not in valid_presets:
raise ValueError(
"Valid presets for class_weight include "
'"balanced" and "balanced_subsample".'
'Given "%s".' % self.class_weight
)
if self.warm_start:
warn(
'class_weight presets "balanced" or '
'"balanced_subsample" are '
"not recommended for warm_start if the fitted data "
"differs from the full dataset. In order to use "
'"balanced" weights, use compute_class_weight '
'("balanced", classes, y). In place of y you can use '
"a large enough sample of the full training set "
"target to properly estimate the class frequency "
"distributions. Pass the resulting weights as the "
"class_weight parameter.",
obj=self,
)
if self.class_weight != "balanced_subsample" or not self.bootstrap:
if self.class_weight == "balanced_subsample":
class_weight = "balanced"
else:
class_weight = self.class_weight
expanded_class_weight = compute_sample_weight(class_weight, y_original)
return y, expanded_class_weight
@classmethod
def get_test_params(cls, parameter_set="default"):
"""Return testing parameter settings for the estimator.
Parameters
----------
parameter_set : str, default="default"
Name of the set of test parameters to return, for use in tests. If no
special parameters are defined for a value, will return ``"default"`` set.
For classifiers, a "default" set of parameters should be provided for
general testing, and a "results_comparison" set for comparing against
previously recorded results if the general set does not produce suitable
probabilities to compare against.
Returns
-------
params : dict or list of dict, default={}
Parameters to create testing instances of the class.
Each dict are parameters to construct an "interesting" test instance, i.e.,
``MyClass(**params)`` or ``MyClass(**params[i])`` creates a valid test
instance.
``create_test_instance`` uses the first (or only) dictionary in ``params``.
"""
params1 = {"n_estimators": 2}
params2 = {"n_estimators": 5, "min_samples_split": 3, "bootstrap": True}
return [params1, params2]