-
Notifications
You must be signed in to change notification settings - Fork 89
/
_minirocket_mv.py
1157 lines (1021 loc) · 28.1 KB
/
_minirocket_mv.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
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
"""Multivariate MiniRocket transformer."""
__author__ = ["angus924", "michaelfeil", "TonyBagnall"]
__all__ = ["MiniRocketMultivariateVariable"]
import multiprocessing
import warnings
from typing import List, Union
import numpy as np
import pandas as pd
from numba import get_num_threads, njit, prange, set_num_threads, vectorize
from aeon.transformations.collection import BaseCollectionTransformer
class MiniRocketMultivariateVariable(BaseCollectionTransformer):
"""MINIROCKET (Multivariate, unequal length).
MINImally RandOm Convolutional KErnel Transform. [1]_
**Multivariate** and **unequal length**
A provisional and naive extension of MINIROCKET to multivariate input
with unequal length provided by the authors [2]_ . For better
performance, use the aeon class MiniRocket for univariate input,
and MiniRocketMultivariate to equal length multivariate input.
Parameters
----------
num_kernels : int, default=10,000
number of random convolutional kernels. The calculated number of features is the
nearest multiple of n_features_per_kernel(default 4)*84=336 < 50,000
(2*n_features_per_kernel(default 4)*num_kernels(default 10,000)).
max_dilations_per_kernel : int, default=32
maximum number of dilations per kernel.
reference_length : int or str, default = `'max'`
series-length of reference, str defines how to infer from X during 'fit'.
options are `'max'`, `'mean'`, `'median'`, `'min'`.
pad_value_short_series : float or None, default=None
if padding series with len<9 to value. if None, not padding is performed.
n_jobs : int, default=1
The number of jobs to run in parallel for `transform`. ``-1`` means using all
processors.
random_state : None or int, default = None
Examples
--------
>>> from aeon.transformations.collection.convolution_based import (
... MiniRocketMultivariateVariable
... )
>>> from aeon.datasets import load_japanese_vowels
>>> # load multivariate and unequal length dataset
>>> X_train, _ = load_japanese_vowels(split="train")
>>> X_test, _ = load_japanese_vowels(split="test")
>>> pre_clf = MiniRocketMultivariateVariable(pad_value_short_series=0.0)
>>> pre_clf.fit(X_train, y=None)
MiniRocketMultivariateVariable(pad_value_short_series=0.0)
>>> X_transformed = pre_clf.transform(X_test)
>>> X_transformed.shape
(370, 9996)
Raises
------
ValueError
If any multivariate series_length in X is < 9 and
pad_value_short_series is set to None
See Also
--------
MultiRocket, MiniRocket, MiniRocketMultivariate, Rocket
References
----------
.. [1] Angus Dempster, Daniel F Schmidt, Geoffrey I Webb
MINIROCKET: A Very Fast (Almost) Deterministic Transform for
Time Series Classification, 2020, arXiv:2012.08791
.. [2] Angus Dempster, Daniel F Schmidt, Geoffrey I Webb
https://github.com/angus924/minirocket
"""
_tags = {
"output_data_type": "Tabular",
"capability:multivariate": True,
"capability:unequal_length": True,
"X_inner_type": "np-list",
"algorithm_type": "convolution",
}
def __init__(
self,
num_kernels=10000,
max_dilations_per_kernel=32,
reference_length="max",
pad_value_short_series=None,
n_jobs=1,
random_state=None,
):
self.num_kernels = num_kernels
self.max_dilations_per_kernel = max_dilations_per_kernel
self.reference_length = reference_length
self._fitted_reference_length = None
self.pad_value_short_series = pad_value_short_series
self.n_jobs = n_jobs
self.random_state = random_state
if random_state is None:
self.random_state_ = random_state
elif isinstance(random_state, int):
self.random_state_ = np.int32(random_state)
else:
raise ValueError(
f"random_state in MiniRocketMultivariateVariable must be int or None, "
f"but found <{type(random_state)} {random_state}>"
)
self._reference_modes = ["max", "mean", "median", "min"]
if not (isinstance(reference_length, int) and reference_length >= 9) and not (
isinstance(reference_length, str)
and (reference_length in self._reference_modes)
):
raise ValueError(
"reference_length in MiniRocketMultivariateVariable must be int>=9 or "
"'max', 'mean', 'median', but found reference_length="
f"{reference_length}"
)
super(MiniRocketMultivariateVariable, self).__init__()
def _fit(self, X, y=None):
"""Fits dilations and biases to input time series.
Parameters
----------
X : List of 2D np.ndarray
y : ignored argument for interface compatibility
Returns
-------
self
Raises
------
ValueError
If any multivariate series_length in X is < 9 and
pad_value_short_series is set to None
"""
X_2d_t, lengths_1darray = _np_list_transposed2D_array_and_len_list(
X, pad=self.pad_value_short_series
)
if isinstance(self.reference_length, int):
_reference_length = self.reference_length
elif self.reference_length in self._reference_modes:
# np.mean, np.max, np.median, np.min ..
_reference_length = getattr(np, self.reference_length)(lengths_1darray)
else:
raise ValueError(
"reference_length in MiniRocketMultivariateVariable must be int>=9 or "
"'max', 'mean', 'median', but found reference_length="
f"{self.reference_length}"
)
self._fitted_reference_length = int(max(9, _reference_length))
if lengths_1darray.min() < 9:
failed_index = np.where(lengths_1darray < 9)[0]
raise ValueError(
(
f"X must be >= 9 for all samples, but found miniumum to be "
f"{lengths_1darray.min()}; at index {failed_index}, pad shorter "
"series so that n_timepoints >= 9 for all samples."
)
)
if lengths_1darray.min() == lengths_1darray.max():
warnings.warn(
"X is of equal length, consider using MiniRocketMultivariate for "
"speedup and stability instead.",
stacklevel=2,
)
if X_2d_t.shape[0] == 1:
warnings.warn(
"X is univariate, consider using the univariate MiniRocket for "
"speedup and stability instead.",
stacklevel=2,
)
self.parameters = _fit_multi_var(
X_2d_t,
L=lengths_1darray,
reference_length=self._fitted_reference_length,
num_features=self.num_kernels,
max_dilations_per_kernel=self.max_dilations_per_kernel,
seed=self.random_state_,
)
return self
def _transform(self, X, y=None):
"""Transform input time series.
Parameters
----------
X : 2D list on np.ndarray
y : ignored argument for interface compatibility
Returns
-------
np.ndarray, size (n_instances, num_kernels)
Raises
------
ValueError
If any multivariate series_length in X is < 9 and
pad_value_short_series is set to None
"""
X_2d_t, L = _np_list_transposed2D_array_and_len_list(
X, pad=self.pad_value_short_series
)
# change n_jobs dependend on value and existing cores
prev_threads = get_num_threads()
if self.n_jobs < 1 or self.n_jobs > multiprocessing.cpu_count():
n_jobs = multiprocessing.cpu_count()
else:
n_jobs = self.n_jobs
set_num_threads(n_jobs)
X_ = _transform_multi_var(X_2d_t, L, self.parameters)
set_num_threads(prev_threads)
return X_
def _np_list_transposed2D_array_and_len_list(
X: List[pd.DataFrame], pad: Union[int, float, None] = 0
):
"""Convert a list of 2D numpy to a 2D array and a list of lengths.
Parameters
----------
X : List of dataframes
List of length n_instances, with
dataframes of series_length-rows and n_channels-columns
pad : float or None. if float/int,pads multivariate series with 'pad',
so that each series has at least length 9.
if None, no padding is applied.
Returns
-------
np.array: 2D array of shape =
[n_channels, sum(length_series(i) for i in n_instances)],
np.float32
np.array: 1D array of shape = [n_instances]
with length of each series, np.int32
Raises
------
ValueError
If any multivariate series_length in X is < 9 and
pad_value_short_series is set to None
"""
vec = []
lengths = []
for _x in X:
_x_shape = _x.shape
if _x_shape[1] < 9:
if pad is not None:
# emergency: pad with zeros up to 9.
lengths.append(9)
padding_width = ((0, 0), (0, 9 - _x_shape[1]))
x_pad = np.pad(_x, padding_width, mode="constant", constant_values=pad)
vec.append(np.transpose(x_pad))
else:
raise ValueError(
"X series_length must be >= 9 for all samples"
f"but sample with series_length {_x_shape[1]} found. Consider"
" padding, discard, or setting a pad_value_short_series value"
)
else:
lengths.append(_x_shape[1])
vec.append(np.transpose(_x))
X_2d_t = np.vstack(vec).T.astype(dtype=np.float32)
lengths = np.array(lengths, dtype=np.int32)
if not lengths.sum() == X_2d_t.shape[1]:
raise ValueError("X_new and lengths do not match. check input dimension")
return X_2d_t, lengths
# code below from the orignal authors: https://github.com/angus924/minirocket
@njit(
"float32[:](float32[:,:],int32[:],int32[:],int32[:],int32[:],int32[:],float32[:],"
"optional(int32))",
fastmath=True,
parallel=False,
cache=True,
)
def _fit_biases_multi_var(
X,
L,
num_channels_per_combination,
channel_indices,
dilations,
num_features_per_dilation,
quantiles,
seed,
):
if seed is not None:
np.random.seed(seed)
n_instances = len(L)
num_channels, _ = X.shape
# equivalent to:
# >>> from itertools import combinations
# >>> indices = np.array(
# >>> [_ for _ in combinations(np.arange(9), 3)], dtype = np.int32
# >>> )
indices = np.array(
(
0,
1,
2,
0,
1,
3,
0,
1,
4,
0,
1,
5,
0,
1,
6,
0,
1,
7,
0,
1,
8,
0,
2,
3,
0,
2,
4,
0,
2,
5,
0,
2,
6,
0,
2,
7,
0,
2,
8,
0,
3,
4,
0,
3,
5,
0,
3,
6,
0,
3,
7,
0,
3,
8,
0,
4,
5,
0,
4,
6,
0,
4,
7,
0,
4,
8,
0,
5,
6,
0,
5,
7,
0,
5,
8,
0,
6,
7,
0,
6,
8,
0,
7,
8,
1,
2,
3,
1,
2,
4,
1,
2,
5,
1,
2,
6,
1,
2,
7,
1,
2,
8,
1,
3,
4,
1,
3,
5,
1,
3,
6,
1,
3,
7,
1,
3,
8,
1,
4,
5,
1,
4,
6,
1,
4,
7,
1,
4,
8,
1,
5,
6,
1,
5,
7,
1,
5,
8,
1,
6,
7,
1,
6,
8,
1,
7,
8,
2,
3,
4,
2,
3,
5,
2,
3,
6,
2,
3,
7,
2,
3,
8,
2,
4,
5,
2,
4,
6,
2,
4,
7,
2,
4,
8,
2,
5,
6,
2,
5,
7,
2,
5,
8,
2,
6,
7,
2,
6,
8,
2,
7,
8,
3,
4,
5,
3,
4,
6,
3,
4,
7,
3,
4,
8,
3,
5,
6,
3,
5,
7,
3,
5,
8,
3,
6,
7,
3,
6,
8,
3,
7,
8,
4,
5,
6,
4,
5,
7,
4,
5,
8,
4,
6,
7,
4,
6,
8,
4,
7,
8,
5,
6,
7,
5,
6,
8,
5,
7,
8,
6,
7,
8,
),
dtype=np.int32,
).reshape(84, 3)
num_kernels = len(indices)
num_dilations = len(dilations)
num_features = num_kernels * np.sum(num_features_per_dilation)
biases = np.zeros(num_features, dtype=np.float32)
feature_index_start = 0
combination_index = 0
num_channels_start = 0
for dilation_index in range(num_dilations):
dilation = dilations[dilation_index]
padding = ((9 - 1) * dilation) // 2
num_features_this_dilation = num_features_per_dilation[dilation_index]
for kernel_index in range(num_kernels):
feature_index_end = feature_index_start + num_features_this_dilation
num_channels_this_combination = num_channels_per_combination[
combination_index
]
num_channels_end = num_channels_start + num_channels_this_combination
channels_this_combination = channel_indices[
num_channels_start:num_channels_end
]
example_index = np.random.randint(n_instances)
input_length = np.int64(L[example_index])
b = np.sum(L[0 : example_index + 1])
a = b - input_length
_X = X[channels_this_combination, a:b]
A = -_X # A = alpha * X = -X
G = _X + _X + _X # G = gamma * X = 3X
C_alpha = np.zeros(
(num_channels_this_combination, input_length), dtype=np.float32
)
C_alpha[:] = A
C_gamma = np.zeros(
(9, num_channels_this_combination, input_length), dtype=np.float32
)
C_gamma[9 // 2] = G
start = dilation
end = input_length - padding
for gamma_index in range(9 // 2):
# thanks to Murtaza Jafferji @murtazajafferji for suggesting this fix
if end > 0:
C_alpha[:, -end:] = C_alpha[:, -end:] + A[:, :end]
C_gamma[gamma_index, :, -end:] = G[:, :end]
end += dilation
for gamma_index in range(9 // 2 + 1, 9):
if start < input_length:
C_alpha[:, :-start] = C_alpha[:, :-start] + A[:, start:]
C_gamma[gamma_index, :, :-start] = G[:, start:]
start += dilation
index_0, index_1, index_2 = indices[kernel_index]
C = C_alpha + C_gamma[index_0] + C_gamma[index_1] + C_gamma[index_2]
C = np.sum(C, axis=0)
biases[feature_index_start:feature_index_end] = np.quantile(
C, quantiles[feature_index_start:feature_index_end]
)
feature_index_start = feature_index_end
combination_index += 1
num_channels_start = num_channels_end
return biases
def _fit_dilations_multi_var(reference_length, num_features, max_dilations_per_kernel):
num_kernels = 84
num_features_per_kernel = num_features // num_kernels
true_max_dilations_per_kernel = min(
num_features_per_kernel, max_dilations_per_kernel
)
multiplier = num_features_per_kernel / true_max_dilations_per_kernel
max_exponent = np.log2((reference_length - 1) / (9 - 1))
dilations, num_features_per_dilation = np.unique(
np.logspace(0, max_exponent, true_max_dilations_per_kernel, base=2).astype(
np.int32
),
return_counts=True,
)
num_features_per_dilation = (num_features_per_dilation * multiplier).astype(
np.int32
) # this is a vector
remainder = num_features_per_kernel - np.sum(num_features_per_dilation)
i = 0
while remainder > 0:
num_features_per_dilation[i] += 1
remainder -= 1
i = (i + 1) % len(num_features_per_dilation)
return dilations, num_features_per_dilation
# low-discrepancy sequence to assign quantiles to kernel/dilation combinations
def _quantiles_multi_var(n):
return np.array(
[(_ * ((np.sqrt(5) + 1) / 2)) % 1 for _ in range(1, n + 1)], dtype=np.float32
)
def _fit_multi_var(
X,
L,
reference_length: int,
num_features=10_000,
max_dilations_per_kernel=32,
seed=None,
):
if seed is not None:
np.random.seed(seed)
# note in relation to dilation:
# * change *reference_length* according to what is appropriate for your
# application, e.g., L.max(), L.mean(), np.median(L)
# * use _fit_multi_var(...) with an appropriate subset of time series, e.g., for
# reference_length = L.mean(), call _fit_multi_var(...) using only time series
# of at least length L.mean() [see filter_by_length(...)]
if reference_length is None:
raise ValueError("reference_length must be specified")
num_channels, _ = X.shape
num_kernels = 84
dilations, num_features_per_dilation = _fit_dilations_multi_var(
reference_length, num_features, max_dilations_per_kernel
)
num_features_per_kernel = np.sum(num_features_per_dilation)
quantiles = _quantiles_multi_var(num_kernels * num_features_per_kernel)
num_dilations = len(dilations)
num_combinations = num_kernels * num_dilations
max_num_channels = min(num_channels, 9)
max_exponent = np.log2(max_num_channels + 1)
num_channels_per_combination = (
2 ** np.random.uniform(0, max_exponent, num_combinations)
).astype(np.int32)
channel_indices = np.zeros(num_channels_per_combination.sum(), dtype=np.int32)
num_channels_start = 0
for combination_index in range(num_combinations):
num_channels_this_combination = num_channels_per_combination[combination_index]
num_channels_end = num_channels_start + num_channels_this_combination
channel_indices[num_channels_start:num_channels_end] = np.random.choice(
num_channels, num_channels_this_combination, replace=False
)
num_channels_start = num_channels_end
biases = _fit_biases_multi_var(
X,
L,
num_channels_per_combination,
channel_indices,
dilations,
num_features_per_dilation,
quantiles,
seed,
)
return (
num_channels_per_combination,
channel_indices,
dilations,
num_features_per_dilation,
biases,
)
@vectorize("float32(float32,float32)", nopython=True, cache=True)
def _PPV(a, b):
if a > b:
return 1
else:
return 0
@njit(
"float32[:,:](float32[:,:],int32[:],Tuple((int32[:],int32[:],int32[:],int32[:],"
"float32[:])))",
fastmath=True,
parallel=True,
cache=True,
)
def _transform_multi_var(X, L, parameters):
n_instances = len(L)
num_channels, _ = X.shape
(
num_channels_per_combination,
channel_indices,
dilations,
num_features_per_dilation,
biases,
) = parameters
# equivalent to:
# >>> from itertools import combinations
# >>> indices = np.array(
# >>> [_ for _ in combinations(np.arange(9), 3)], dtype = np.int32
# >>> )
indices = np.array(
(
0,
1,
2,
0,
1,
3,
0,
1,
4,
0,
1,
5,
0,
1,
6,
0,
1,
7,
0,
1,
8,
0,
2,
3,
0,
2,
4,
0,
2,
5,
0,
2,
6,
0,
2,
7,
0,
2,
8,
0,
3,
4,
0,
3,
5,
0,
3,
6,
0,
3,
7,
0,
3,
8,
0,
4,
5,
0,
4,
6,
0,
4,
7,
0,
4,
8,
0,
5,
6,
0,
5,
7,
0,
5,
8,
0,
6,
7,
0,
6,
8,
0,
7,
8,
1,
2,
3,
1,
2,
4,
1,
2,
5,
1,
2,
6,
1,
2,
7,
1,
2,
8,
1,
3,
4,
1,
3,
5,
1,
3,
6,
1,
3,
7,
1,
3,
8,
1,
4,
5,
1,
4,
6,
1,
4,
7,
1,
4,
8,
1,
5,
6,
1,
5,
7,
1,
5,
8,
1,
6,
7,
1,
6,
8,
1,
7,
8,
2,
3,
4,
2,
3,
5,
2,
3,
6,
2,
3,
7,
2,
3,
8,
2,
4,
5,
2,
4,
6,
2,
4,
7,
2,
4,
8,
2,
5,
6,
2,
5,
7,
2,
5,
8,
2,
6,
7,