Repository navigation
Expand file tree
/
Copy pathtest_movestrategy.py
More file actions
1436 lines (1252 loc) · 59 KB
/
Copy pathtest_movestrategy.py
File metadata and controls
1436 lines (1252 loc) · 59 KB
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
from __future__ import division
from __future__ import print_function
from __future__ import absolute_import
from builtins import zip
from builtins import range
from past.utils import old_div
from builtins import object
from numpy.testing import assert_almost_equal
from .test_helpers import (
true_func, assert_equal_array_array, make_1d_traj, MoverWithSignature,
setify_ensemble_signature, reorder_ensemble_signature
)
import pytest
import openpathsampling as paths
from openpathsampling.high_level.move_scheme import MoveScheme, DefaultScheme
from openpathsampling.high_level.move_strategy import *
import collections
import logging
logging.getLogger('openpathsampling.initialization').setLevel(logging.CRITICAL)
logging.getLogger('openpathsampling.ensemble').setLevel(logging.CRITICAL)
logging.getLogger('openpathsampling.storage').setLevel(logging.CRITICAL)
logging.getLogger('openpathsampling.netcdfplus').setLevel(logging.CRITICAL)
class MockMoveStrategy(MoveStrategy):
def make_movers(self, scheme):
return None
class MockSingleEnsembleMoveStrategy(MockMoveStrategy,
SingleEnsembleMoveStrategy):
pass
def find_mover(scheme, group, sig):
mover = None
for m in scheme.movers[group]:
if m.ensemble_signature_set == (set(sig[0]), set(sig[1])):
mover = m
return mover
class TestStrategyLevels(object):
def test_level_type(self):
assert levels.level_type(10) == levels.SIGNATURE
assert levels.level_type(1) == levels.SIGNATURE
assert levels.level_type(19) == levels.SIGNATURE
assert levels.level_type(20) is None
assert levels.level_type(21) == levels.MOVER
assert levels.level_type(35) == levels.MOVER
assert levels.level_type(100) == levels.GLOBAL
class MoveStrategyTestSetup(object):
def setup_method(self):
paths.InterfaceSet._reset()
cvA = paths.FunctionCV(name="xA", f=lambda s: s.xyz[0][0])
cvB = paths.FunctionCV(name="xB", f=lambda s: -s.xyz[0][0])
self.stateA = paths.CVDefinedVolume(cvA, float("-inf"), -0.5)
self.stateB = paths.CVDefinedVolume(cvB, float("-inf"), -0.5)
interfacesA = paths.VolumeInterfaceSet(cvA, float("-inf"),
[-0.5, -0.3, -0.1])
interfacesB = paths.VolumeInterfaceSet(cvB, float("-inf"),
[-0.5, -0.3, -0.1])
self.network = paths.MSTISNetwork(
[(self.stateA, interfacesA),
(self.stateB, interfacesB)],
ms_outers=paths.MSOuterTISInterface.from_lambdas(
{interfacesA: 0.0, interfacesB: 0.0}
)
)
class TestMoveStrategy(MoveStrategyTestSetup):
def test_levels(self):
strategy = MockMoveStrategy(ensembles=None, group="test", replace=True)
assert strategy.level == -1
assert strategy.replace_signatures is False
assert strategy.replace_movers is False
strategy.level = 10
assert strategy.level == levels.SIGNATURE
assert strategy.replace_signatures
assert not strategy.replace_movers
strategy.level = 25
assert strategy.level != levels.MOVER
assert levels.level_type(strategy.level) == levels.MOVER
assert not strategy.replace_signatures
assert strategy.replace_movers
strategy.level = 99
assert not strategy.replace_signatures
assert not strategy.replace_movers
def test_get_ensembles(self):
self.strategy = MockMoveStrategy(ensembles=None, group="test",
replace=True)
scheme = MoveScheme(self.network)
# load up the relevant ensembles to test against
transition_ensembles = []
for transition in self.network.sampling_transitions:
transition_ensembles.append(transition.ensembles)
assert len(transition_ensembles) == 2
for ens_set in transition_ensembles:
assert len(ens_set) == 3
ensA = self.network.from_state[self.stateA].ensembles
assert len(ensA) == 3
# if you error before this, something is wrong in setup
ensembles = self.strategy.get_ensembles(scheme, None)
assert ensembles == transition_ensembles
ensembles = self.strategy.get_ensembles(scheme, ensA)
assert ensembles == [ensA]
extra_ens = transition_ensembles[1][0]
weird_ens_list = [[ensA[0]], ensA[1], [extra_ens]]
ensembles = self.strategy.get_ensembles(scheme, weird_ens_list)
assert ensembles == [[ensA[0]], [ensA[1]], [extra_ens]]
ensembles = self.strategy.get_ensembles(scheme, extra_ens)
assert len(ensembles) == 1
assert len(ensembles[0]) == 1
assert ensembles[0][0] == extra_ens
class TestSingleEnsembleMoveStrategy(MoveStrategyTestSetup):
def setup_method(self):
super(TestSingleEnsembleMoveStrategy, self).setup_method()
self.strategy = MockSingleEnsembleMoveStrategy(
ensembles=None,
group='test_group',
replace=True
)
self.scheme = paths.DefaultScheme(self.network, engine=None)
def test_get_per_mover_ensembles(self):
per_mover_ensembles = \
self.strategy.get_per_mover_ensembles(self.scheme)
ensembles = self.scheme.network.sampling_ensembles
assert len(per_mover_ensembles) == len(ensembles)
for listed_ensemble in per_mover_ensembles:
assert len(listed_ensemble) == 1
assert listed_ensemble[0] in ensembles
def test_get_parameters(self):
list_params_as_list = [0, 1, 2, 3, 4, 5]
list_params_as_single = 100
nonlist_params = "alpha"
params = self.strategy.get_parameters(
scheme=self.scheme,
list_parameters=[list_params_as_list, list_params_as_single],
nonlist_parameters=[nonlist_params]
)
ensembles = self.scheme.network.sampling_ensembles
expected = [(ens, num, list_params_as_single, nonlist_params)
for (ens, num) in zip(ensembles, list_params_as_list)]
assert params == expected
def test_get_parameters_error(self):
with pytest.raises(RuntimeError):
params = self.strategy.get_parameters(
scheme=self.scheme,
list_parameters=[[0, 1, 2, 3]]
)
class TestForwardShootingStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = ForwardShootingStrategy()
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 6
for mover in movers:
assert type(mover) == paths.ForwardShootMover
assert type(mover.selector) == paths.UniformSelector
def test_make_movers_with_list(self):
list_of_selectors = [
paths.shooting.InterfaceConstrainedSelector(ens.interface)
for ens in self.network.sampling_ensembles
]
strategy = ForwardShootingStrategy(
selector=list_of_selectors,
ensembles=self.network.sampling_ensembles
)
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 6
assert len(list_of_selectors) == 6
for mover, sel in zip(movers, list_of_selectors):
assert type(mover) == paths.ForwardShootMover
assert (type(mover.selector)
== paths.shooting.InterfaceConstrainedSelector)
assert mover.selector == sel
class TestOneWayShootingStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = OneWayShootingStrategy()
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 6
for mover in movers:
assert type(mover) == paths.OneWayShootingMover
assert type(mover.selector) == paths.UniformSelector
class TestTwoWayShootingStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = TwoWayShootingStrategy(modifier=paths.NoModification())
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 6
for mover in movers:
assert type(mover) == paths.TwoWayShootingMover
assert type(mover.selector) == paths.UniformSelector
assert type(mover.modifier) == paths.NoModification
def test_composition_with_default_scheme(self):
strategy = TwoWayShootingStrategy(modifier=paths.NoModification())
scheme = DefaultScheme(self.network, engine=None)
scheme.append(strategy)
scheme.build_move_decision_tree()
assert len(scheme.movers['shooting']) == 6
for mover in scheme.movers['shooting']:
assert type(mover) == paths.TwoWayShootingMover
assert type(mover.selector) == paths.UniformSelector
assert type(mover.modifier) == paths.NoModification
class TestNearestNeighborRepExStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = NearestNeighborRepExStrategy()
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 4
ens0 = self.network.sampling_transitions[0].ensembles
ens1 = self.network.sampling_transitions[1].ensembles
assert (movers[0].ensemble_signature_set
== (set([ens0[0], ens0[1]]), set([ens0[0], ens0[1]])))
assert (movers[1].ensemble_signature_set
== (set([ens0[1], ens0[2]]), set([ens0[1], ens0[2]])))
assert (movers[2].ensemble_signature_set
== (set([ens1[0], ens1[1]]), set([ens1[0], ens1[1]])))
assert (movers[3].ensemble_signature_set
== (set([ens1[1], ens1[2]]), set([ens1[1], ens1[2]])))
class TestAllSetRepExStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = AllSetRepExStrategy()
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 6
ens0 = self.network.sampling_transitions[0].ensembles
ens1 = self.network.sampling_transitions[1].ensembles
signatures = [(set(m.ensemble_signature[0]),
set(m.ensemble_signature[1])) for m in movers]
expected_signatures = [
((ens0[0], ens0[1]), (ens0[0], ens0[1])),
((ens0[0], ens0[2]), (ens0[0], ens0[2])),
((ens0[1], ens0[2]), (ens0[1], ens0[2])),
((ens1[0], ens1[1]), (ens1[0], ens1[1])),
((ens1[0], ens1[2]), (ens1[0], ens1[2])),
((ens1[1], ens1[2]), (ens1[1], ens1[2]))
]
for sig in expected_signatures:
set_sig = (set(sig[0]), set(sig[1]))
assert set_sig in signatures
class TestSelectedPairsRepExStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
ens00 = self.network.sampling_transitions[0].ensembles[0]
ens02 = self.network.sampling_transitions[0].ensembles[2]
strategy = SelectedPairsRepExStrategy(ensembles=[ens00, ens02])
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 1
assert (movers[0].ensemble_signature_set
== ({ ens00, ens02 }, ({ ens00, ens02 })))
def test_init_ensembles_none(self):
with pytest.raises(RuntimeError):
strategy = SelectedPairsRepExStrategy()
def test_init_ensembles_triplet(self):
ensembles = self.network.sampling_transitions[0].ensembles
with pytest.raises(RuntimeError):
strategy = SelectedPairsRepExStrategy(ensembles=ensembles)
def test_make_movers_multiple_pairs(self):
ens00 = self.network.sampling_transitions[0].ensembles[0]
ens01 = self.network.sampling_transitions[0].ensembles[1]
ens02 = self.network.sampling_transitions[0].ensembles[2]
strategy = SelectedPairsRepExStrategy(ensembles=[[ens00, ens01],
[ens00, ens02],
[ens01, ens02]])
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 3
assert (movers[0].ensemble_signature_set
== ({ens00, ens01}, {ens00, ens01}))
assert (movers[1].ensemble_signature_set
== ({ens00, ens02}, {ens00, ens02}))
assert (movers[2].ensemble_signature_set
== ({ens01, ens02}, {ens01, ens02}))
class TestReplicaExchangeStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = ReplicaExchangeStrategy()
scheme = MoveScheme(self.network)
scheme.apply_strategy(NearestNeighborRepExStrategy())
movers = strategy.make_movers(scheme)
assert len(movers) == 4
def test_swap_to_hop_to_swap(self):
scheme = DefaultScheme(self.network)
root = scheme.move_decision_tree()
assert len(scheme.movers['repex']) == 6
old_movers = scheme.movers['repex']
scheme.append(EnsembleHopStrategy(), force=True)
root = scheme.move_decision_tree(rebuild=True)
assert len(scheme.movers['repex']) == 12
scheme.append(ReplicaExchangeStrategy(), force=True)
root = scheme.move_decision_tree(rebuild=True)
assert len(scheme.movers['repex']) == 6
new_movers = scheme.movers['repex']
assert old_movers != new_movers
old_sigs = [m.ensemble_signature_set for m in old_movers]
new_sigs = [m.ensemble_signature_set for m in new_movers]
for old in old_sigs:
assert old in new_sigs
def test_detailed_balance_partners(self):
scheme = DefaultScheme(self.network)
scheme.append(EnsembleHopStrategy(), force=True)
root = scheme.move_decision_tree()
assert len(scheme.movers['repex']) == 12
scheme.movers['repex'].pop()
assert len(scheme.movers['repex']) == 11
strategy = ReplicaExchangeStrategy()
with pytest.raises(RuntimeError):
strategy.make_movers(scheme)
class TestEnsembleHopStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = EnsembleHopStrategy()
scheme = MoveScheme(self.network)
scheme.apply_strategy(NearestNeighborRepExStrategy())
movers = strategy.make_movers(scheme)
# defaults to 4 repex movers, so
assert len(movers) == 8
# set up the swap pairs
swap_pairs = []
for trans in self.network.sampling_transitions:
swap_pairs.extend([[trans.ensembles[i], trans.ensembles[i+1]]
for i in range(len(trans.ensembles)-1)])
assert len(swap_pairs) == 4
# check that each swap pair has a hop
mover_sigs = [m.ensemble_signature for m in movers]
for pair in swap_pairs:
sig1 = ((pair[0],),(pair[1],))
sig2 = ((pair[1],),(pair[0],))
assert sig1 in mover_sigs
assert sig2 in mover_sigs
scheme.movers['repex'] = movers
newmovers = strategy.make_movers(scheme)
assert len(newmovers) == 8
for mover in newmovers:
assert mover.ensemble_signature in mover_sigs
def test_different_number_input_output_ensembles(self):
ens0 = self.network.sampling_ensembles[0]
ens1 = self.network.sampling_ensembles[1]
ens2 = self.network.sampling_ensembles[2]
weird_mover = MoverWithSignature(
input_ensembles=[ens0, ens1, ens2],
output_ensembles=[ens0, ens1]
)
assert (weird_mover.ensemble_signature
== ((ens0,ens1,ens2),(ens0,ens1)))
scheme = MoveScheme(self.network)
scheme.movers['weird'] = [weird_mover]
strategy = EnsembleHopStrategy(group='weird')
with pytest.raises(RuntimeError):
strategy.make_movers(scheme)
def test_wrong_number_ensembles_in_signature(self):
ens0 = self.network.sampling_ensembles[0]
ens1 = self.network.sampling_ensembles[1]
ens2 = self.network.sampling_ensembles[2]
weird_mover = MoverWithSignature(
input_ensembles=[ens0, ens1, ens2],
output_ensembles=[ens0, ens1, ens2]
)
assert (weird_mover.ensemble_signature
== ((ens0,ens1,ens2),(ens0,ens1,ens2)))
scheme = MoveScheme(self.network)
scheme.movers['weird'] = [weird_mover]
strategy = EnsembleHopStrategy(group='weird')
with pytest.raises(RuntimeError):
strategy.make_movers(scheme)
def test_not_replica_exchange_signature(self):
ens0 = self.network.sampling_ensembles[0]
ens1 = self.network.sampling_ensembles[1]
ens2 = self.network.sampling_ensembles[2]
weird_mover = MoverWithSignature(
input_ensembles=[ens0, ens1],
output_ensembles=[ens1, ens2]
)
assert weird_mover.ensemble_signature == ((ens0,ens1),(ens1,ens2))
scheme = MoveScheme(self.network)
scheme.movers['weird'] = [weird_mover]
strategy = EnsembleHopStrategy(group='weird')
with pytest.raises(RuntimeError):
strategy.make_movers(scheme)
def test_replace_nofrom(self):
# this is the default behavior
scheme = DefaultScheme(self.network)
scheme.movers ={}
scheme.append(EnsembleHopStrategy(replace=True, from_group=None))
scheme.build_move_decision_tree()
# 4 normal repex + 2 ms-outer repex = 6 repex * 2 hop/repex = 12
assert len(scheme.movers['repex']) == 12
def test_noreplace_from(self):
# if replace is False and from_group is given, we end up with two
# groups: both the new and the old.
scheme = DefaultScheme(self.network)
scheme.movers ={}
scheme.append(EnsembleHopStrategy(replace=False,
group='hop',
from_group='repex'))
scheme.build_move_decision_tree()
assert len(scheme.movers['repex']) == 6
assert len(scheme.movers['hop']) == 12
def test_replace_from(self):
# if replace is True and we have a different from_group, we should
# remove the old from_group from existence
scheme = DefaultScheme(self.network)
scheme.movers ={}
scheme.append(EnsembleHopStrategy(replace=True,
group='hop',
from_group='repex'))
scheme.build_move_decision_tree()
assert len(scheme.movers['hop']) == 12
assert "repex" not in list(scheme.movers.keys())
def test_noreplace_nofrom(self):
# if replace==False and the from_group is not given, we should
# act as mover replacement (but this is seriously a bad idea)
scheme = DefaultScheme(self.network)
scheme.movers ={}
scheme.append(EnsembleHopStrategy(replace=False, from_group=None))
scheme.build_move_decision_tree()
assert len(scheme.movers['repex']) == 18
class TestPathReversalStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = PathReversalStrategy()
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 6
for m in movers:
assert type(m) == paths.PathReversalMover
class TestMinusMoveStrategy(MoveStrategyTestSetup):
def test_get_ensembles(self):
strategy = MinusMoveStrategy()
scheme = MoveScheme(self.network)
ensembles = strategy.get_ensembles(scheme, None)
assert len(ensembles) == 2
for ens_group in ensembles:
assert len(ens_group) == 1
assert ensembles[0][0].state_vol != ensembles[1][0].state_vol
def test_get_ensembles_multiple_minus(self):
strategy = MinusMoveStrategy()
innerA = self.network.sampling_transitions[0].ensembles[0]
innerB = self.network.sampling_transitions[1].ensembles[0]
inner_vol_A = self.network.sampling_transitions[0].interfaces[0]
inner_vol_B = self.network.sampling_transitions[1].interfaces[0]
extra_minus = paths.MinusInterfaceEnsemble(
state_vol=self.network.sampling_transitions[0].stateA,
innermost_vols=[inner_vol_A, inner_vol_B]
)
self.network.special_ensembles['minus'][extra_minus] = [innerA, innerB]
scheme = MoveScheme(self.network)
ensembles = strategy.get_ensembles(scheme, None)
assert len(ensembles) == 2
assert { len(ensembles[0]), len(ensembles[1]) } == { 1, 2 }
def test_get_ensembles_fixed_ensembles(self):
strategy = MinusMoveStrategy()
minusA = list(self.network.special_ensembles['minus'].keys())[0]
scheme = MoveScheme(self.network)
ensembles = strategy.get_ensembles(scheme, minusA)
assert len(ensembles) == 1
assert len(ensembles[0]) == 1
assert ensembles[0][0] == minusA
def test_make_movers(self):
strategy = MinusMoveStrategy()
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 2
minuses = self.network.special_ensembles['minus']
ens_minusA = list(minuses.keys())[0]
ens_innerA = [t.ensembles[0] for t in minuses[ens_minusA]]
sig_A = set([ens_minusA] + ens_innerA)
ens_minusB = list(minuses.keys())[1]
ens_innerB = [t.ensembles[0] for t in minuses[ens_minusB]]
sig_B = set([ens_minusB] + ens_innerB)
all_ens_sigs = [m.ensemble_signature_set for m in movers]
# check the signatures
assert sig_A != sig_B
assert tuple([sig_A, sig_A]) in all_ens_sigs
assert tuple([sig_B, sig_B]) in all_ens_sigs
# check that these are inner ensembles
inners = [t.ensembles[0] for t in self.network.sampling_transitions]
for inner in ens_innerA + ens_innerB:
assert inner in inners
# check that we've got the right inner for the right state
stateA_inner = self.network.from_state[ens_minusA.state_vol].ensembles[0]
assert [stateA_inner] == ens_innerA
stateB_inner = self.network.from_state[ens_minusB.state_vol].ensembles[0]
assert [stateB_inner] == ens_innerB
# check that we've got minus ensembles
for mover in movers:
assert mover.minus_ensemble in self.network.minus_ensembles
assert isinstance(mover.minus_ensemble,
paths.MinusInterfaceEnsemble)
class TestSingleReplicaMinusMoveStrategy(MoveStrategyTestSetup):
def test_make_movers(self):
strategy = SingleReplicaMinusMoveStrategy()
scheme = MoveScheme(self.network)
movers = strategy.make_movers(scheme)
assert len(movers) == 2
minuses = self.network.special_ensembles['minus']
ens_minusA = list(minuses.keys())[0]
ens_innerA = [t.ensembles[0] for t in minuses[ens_minusA]]
sig_A = set([ens_minusA] + ens_innerA)
ens_minusB = list(minuses.keys())[1]
ens_innerB = [t.ensembles[0] for t in minuses[ens_minusB]]
sig_B = set([ens_minusB] + ens_innerB)
all_ens_sigs = [m.ensemble_signature_set for m in movers]
class TestOrganizeByMoveGroupStrategy(MoveStrategyTestSetup):
def scheme_setup_shooting_repex(self):
scheme = MoveScheme(self.network)
ens0 = self.network.sampling_transitions[0].ensembles[0]
ens1 = self.network.sampling_transitions[0].ensembles[1]
ens2 = self.network.sampling_transitions[0].ensembles[2]
scheme.movers['shooting'] = [
paths.OneWayShootingMover(
selector=paths.UniformSelector(),
ensemble=ens
)
for ens in [ens0, ens1, ens2]
]
scheme.movers['repex'] = [
paths.ReplicaExchangeMover(ens0, ens1),
paths.ReplicaExchangeMover(ens1, ens2)
]
return scheme
def test_choice_probability(self):
scheme = self.scheme_setup_shooting_repex()
ens0 = self.network.sampling_transitions[0].ensembles[0]
strategy = OrganizeByMoveGroupStrategy()
group_weights = {'shooting' : 1.0, 'repex' : 0.5}
mover_weights = {}
for groupname in list(scheme.movers.keys()):
for mover in scheme.movers[groupname]:
mover_weights[(groupname, mover.ensemble_signature)] = 1.0
choice_prob = strategy.choice_probability(scheme, group_weights,
mover_weights)
# shooting moves: P_a*P_b = 1.0*1.0 (3 times)
# repex moves: P_a*P_b = 0.5*1.0 (2 times)
# norm = 3*1.0 + 2*0.5 = 4.0
# each shooting prob: 0.25
# each repex prob: 0.125
assert len(choice_prob) == len(sum(list(scheme.movers.values()), []))
for m in list(choice_prob.keys()):
if m in scheme.movers['shooting']:
assert choice_prob[m] == 0.25
elif m in scheme.movers['repex']:
assert choice_prob[m] == 0.125
else:
raise RuntimeError("Unknown mover "+repr(m))
# now change the choice probability for one of them
ens0_sig = ((ens0,),(ens0,))
mover_weights[('shooting', ens0_sig)] = 2.0
choice_prob = strategy.choice_probability(scheme, group_weights,
mover_weights)
# shooting0: P_a*P_b = 1.0*2.0
# new norm: 2*1.0 + 1*2.0 + 2*0.5 = 5.0
# prob shooting0 : 2.0/5.0 = 0.4
# prob shooting1,2 : 1.0/5.0 = 0.2
# prob repex: 0.5/5.0 = 0.1
for m in list(choice_prob.keys()):
if m in scheme.movers['shooting']:
if m.ensemble_signature == ens0_sig:
assert choice_prob[m] == 0.4
else:
assert choice_prob[m] == 0.2
elif m in scheme.movers['repex']:
assert choice_prob[m] == 0.1
else:
raise RuntimeError("Unknown mover "+repr(m))
def test_choice_probability_bad_group_weights(self):
scheme = MoveScheme(self.network)
ens0 = self.network.sampling_transitions[0].ensembles[0]
ens1 = self.network.sampling_transitions[0].ensembles[1]
ens2 = self.network.sampling_transitions[0].ensembles[2]
scheme.movers['shooting'] = [
paths.OneWayShootingMover(
selector=paths.UniformSelector(),
ensemble=ens
)
for ens in [ens0, ens1, ens2]
]
scheme.movers['repex'] = [
paths.ReplicaExchangeMover(ens0, ens1),
paths.ReplicaExchangeMover(ens1, ens2)
]
strategy = OrganizeByMoveGroupStrategy()
group_weights = {'shooting' : 1.0}
mover_weights = {}
for groupname in list(scheme.movers.keys()):
for mover in scheme.movers[groupname]:
mover_weights[(groupname, mover.ensemble_signature)] = 1.0
assert len(mover_weights) == 5
with pytest.raises(KeyError):
choice_prob = strategy.choice_probability(scheme, group_weights,
mover_weights)
def test_weights_from_choice_probability(self):
scheme = self.scheme_setup_shooting_repex()
ens0 = self.network.sampling_transitions[0].ensembles[0]
strategy = OrganizeByMoveGroupStrategy()
group_weights = {'shooting' : 1.0, 'repex' : 0.5}
mover_weights = {}
for groupname in list(scheme.movers.keys()):
for mover in scheme.movers[groupname]:
mover_weights[(groupname, mover.ensemble_signature)] = 1.0
ens0_sig = ((ens0,),(ens0,))
mover_weights[('shooting', ens0_sig)] = 2.0
choice_prob = strategy.choice_probability(scheme, group_weights,
mover_weights)
(group_w, mover_w) = strategy.weights_from_choice_probability(
scheme, choice_prob
)
assert group_weights == group_w
assert mover_weights == mover_w
#TODO: run more thorough tests of this
def test_chooser_root_weights(self):
scheme = MoveScheme(self.network)
ens0 = self.network.sampling_transitions[0].ensembles[0]
ens1 = self.network.sampling_transitions[0].ensembles[1]
ens2 = self.network.sampling_transitions[0].ensembles[2]
scheme.movers['shooting'] = [
paths.OneWayShootingMover(
selector=paths.UniformSelector(),
ensemble=ens
)
for ens in [ens0, ens1, ens2]
]
scheme.movers['repex'] = [
paths.ReplicaExchangeMover(ens0, ens1),
paths.ReplicaExchangeMover(ens1, ens2)
]
strategy = OrganizeByMoveGroupStrategy()
group_weights = {'shooting' : 1.0, 'repex' : 0.5}
mover_weights = {}
for groupname in list(scheme.movers.keys()):
for mover in scheme.movers[groupname]:
mover_weights[(groupname, mover.ensemble_signature)] = 1.0
w = strategy.chooser_root_weights(scheme, group_weights, mover_weights)
assert w == {'shooting' : 3.0, 'repex' : 1.0}
def test_chooser_mover_weights(self):
scheme = self.scheme_setup_shooting_repex()
strategy = OrganizeByMoveGroupStrategy()
group_weights = {'shooting' : 1.0, 'repex' : 0.5}
mover_weights = {}
for groupname in list(scheme.movers.keys()):
for mover in scheme.movers[groupname]:
mover_weights[(groupname, mover.ensemble_signature)] = 1.0
m_sig = {}
for g in scheme.movers:
for m in scheme.movers[g]:
m_sig[m] = (g, m.ensemble_signature)
for g in scheme.movers:
expected_w = {m : mover_weights[m_sig[m]] for m in scheme.movers[g]}
w = strategy.chooser_mover_weights(scheme, g, mover_weights)
assert w == expected_w
def test_make_movers(self):
scheme = MoveScheme(self.network)
ens0 = self.network.sampling_transitions[0].ensembles[0]
ens1 = self.network.sampling_transitions[0].ensembles[1]
ens2 = self.network.sampling_transitions[0].ensembles[2]
scheme.movers['shooting'] = [
paths.OneWayShootingMover(
selector=paths.UniformSelector(),
ensemble=ens
)
for ens in [ens0, ens1, ens2]
]
scheme.movers['repex'] = [
paths.ReplicaExchangeMover(ensemble1=ens0, ensemble2=ens1),
paths.ReplicaExchangeMover(ensemble1=ens1, ensemble2=ens2)
]
scheme.movers['pathreversal'] = [
paths.PathReversalMover(ensemble=ens)
for ens in [ens0, ens1, ens2]
]
scheme.movers['minus'] = [paths.MinusMover(
minus_ensemble=self.network.minus_ensembles[0],
innermost_ensembles=[ens0]
)]
strategy = OrganizeByMoveGroupStrategy()
root = strategy.make_movers(scheme)
assert len(root.movers) == 4
names = ['ShootingChooser', 'RepexChooser', 'PathreversalChooser',
'MinusChooser']
name_dict = {root.movers[i].name : i for i in range(len(root.movers))}
for name in names:
assert name in list(name_dict.keys())
name = 'ShootingChooser'
weight = root.weights[name_dict[name]]
chooser = root.movers[name_dict[name]]
assert type(chooser) == paths.RandomChoiceMover
assert weight == 3.0
assert len(chooser.movers) == 3
for w in chooser.weights:
assert w == 1.0
name = 'RepexChooser'
weight = root.weights[name_dict[name]]
chooser = root.movers[name_dict[name]]
assert type(chooser) == paths.RandomChoiceMover
assert weight == 1.0
assert len(chooser.movers) == 2
for w in chooser.weights:
assert w == 1.0
name = 'MinusChooser'
weight = root.weights[name_dict[name]]
chooser = root.movers[name_dict[name]]
assert type(chooser) == paths.RandomChoiceMover
assert len(chooser.movers) == 1
assert weight == 0.2
for w in chooser.weights:
assert w == 1.0
def test_make_movers_unknown_group(self):
scheme = MoveScheme(self.network)
ens0 = self.network.sampling_transitions[0].ensembles[0]
ens1 = self.network.sampling_transitions[0].ensembles[1]
ens2 = self.network.sampling_transitions[0].ensembles[2]
scheme.movers['blahblah'] = [
paths.ReplicaExchangeMover(ensemble1=ens0, ensemble2=ens1),
paths.ReplicaExchangeMover(ensemble1=ens1, ensemble2=ens2)
]
strategy = OrganizeByMoveGroupStrategy()
root = strategy.make_movers(scheme)
name_dict = {root.movers[i].name : i for i in range(len(root.movers))}
name = 'BlahblahChooser'
weight = root.weights[name_dict[name]]
chooser = root.movers[name_dict[name]]
assert type(chooser) == paths.RandomChoiceMover
assert weight == 2.0
assert len(chooser.movers) == 2
for w in chooser.weights:
assert w == 1.0
def test_make_movers_custom_group(self):
scheme = MoveScheme(self.network)
ens0 = self.network.sampling_transitions[0].ensembles[0]
ens1 = self.network.sampling_transitions[0].ensembles[1]
ens2 = self.network.sampling_transitions[0].ensembles[2]
scheme.movers['blahblahblah'] = [
paths.ReplicaExchangeMover(ensemble1=ens0, ensemble2=ens1),
paths.ReplicaExchangeMover(ensemble1=ens1, ensemble2=ens2)
]
strategy = OrganizeByMoveGroupStrategy()
strategy.group_weights['blahblahblah'] = 2.0
root = strategy.make_movers(scheme)
name_dict = {root.movers[i].name : i for i in range(len(root.movers))}
name = 'BlahblahblahChooser'
weight = root.weights[name_dict[name]]
chooser = root.movers[name_dict[name]]
assert type(chooser) == paths.RandomChoiceMover
assert weight == 4.0
assert len(chooser.movers) == 2
for w in chooser.weights:
assert w == 1.0
def test_get_weights_scheme_all_unset(self):
strategy = OrganizeByMoveGroupStrategy()
scheme = MoveScheme(self.network)
scheme.append(NearestNeighborRepExStrategy())
scheme.append(OneWayShootingStrategy())
root = scheme.move_decision_tree()
assert len(scheme.movers) == 2
all_movers = scheme.movers['shooting'] + scheme.movers['repex']
all_movers_sigs = [m.ensemble_signature for m in all_movers]
assert strategy.group_weights == {}
assert strategy.mover_weights == {}
(group_weights, mover_weights) = strategy.get_weights(
scheme=scheme,
sorted_movers=scheme.movers
)
assert group_weights == {'shooting' : 1.0, 'repex' : 0.5}
# check that the number of sigs in a each group matches the number
# of movers in that group
for group in list(group_weights.keys()):
mover_sigs = [sig for sig in list(mover_weights.keys())
if sig[0]==group]
assert len(mover_sigs) == len(scheme.movers[group])
for sig in list(mover_weights.keys()):
assert mover_weights[sig] == 1.0
assert sig[1] in all_movers_sigs
# check that we can reuse these in a different scheme
scheme2 = MoveScheme(self.network)
scheme2.append(OneWayShootingStrategy())
root = scheme2.move_decision_tree()
assert len(scheme2.movers) == 1
(group_weights, mover_weights) = strategy.get_weights(
scheme=scheme2,
sorted_movers=scheme2.movers
)
assert group_weights == {'shooting' : 1.0}
for sig in mover_weights:
assert mover_weights[sig] == 1.0
assert sig[1] in [m.ensemble_signature
for m in scheme2.movers[sig[0]]]
for group in list(scheme2.movers.keys()):
mover_sigs = [sig for sig in list(mover_weights.keys())
if sig[0]==group]
assert len(mover_sigs) == len(scheme2.movers[group])
def test_get_weights_both_internal_weights_set(self):
strategy = OrganizeByMoveGroupStrategy()
ensA = self.network.sampling_transitions[0].ensembles[0]
ensA_sig = ((ensA,),(ensA,))
scheme = MoveScheme(self.network)
scheme.append([NearestNeighborRepExStrategy(),
OneWayShootingStrategy(),
strategy])
root = scheme.move_decision_tree()
old_choice_probability = scheme.choice_probability
old_shooter_ensA = [m for m in scheme.movers['shooting']
if m.ensemble_signature == ensA_sig][0]
strategy.group_weights['repex'] = 3.0
strategy.mover_weights[('shooting',ensA_sig)]= 2.0
scheme.choice_probability = {} # this is unset here
(group_weights, mover_weights) = strategy.get_weights(
scheme=scheme,
sorted_movers=scheme.movers,
sort_weights_override=strategy.group_weights,
mover_weights_override=strategy.mover_weights
)
root = scheme.move_decision_tree(rebuild=True)
assert group_weights == {'shooting' : 1.0, 'repex' : 3.0}
expected_mover_weights = {}
for group in scheme.movers:
for mover in scheme.movers[group]:
sig = mover.ensemble_signature
if group == 'shooting' and sig == ensA_sig:
expected_mover_weights[(group,sig)] = 2.0
elif group == 'shooting':
expected_mover_weights[(group,sig)] = 1.0
else:
expected_mover_weights[(group,sig)] = 1.0
assert mover_weights == expected_mover_weights
new_choice_probability = scheme.choice_probability
new_repex_chooser = [m for m in root if m.name=="RepexChooser"][0]
repex_chooser_idx = root.movers.index(new_repex_chooser)
assert (root.weights[repex_chooser_idx]
== 3.0*len(new_repex_chooser.movers))
new_shoot_chooser = [m for m in root if m.name=="ShootingChooser"][0]
new_shooter_ensA = [m for m in scheme.movers['shooting']
if m.ensemble_signature == ensA_sig][0]
shooter_ensA_idx = new_shoot_chooser.movers.index(new_shooter_ensA)
assert new_shoot_chooser.weights[shooter_ensA_idx] == 2.0
assert new_choice_probability != old_choice_probability
def test_get_weights_group_weights_set(self):
strategy = OrganizeByMoveGroupStrategy()
scheme = MoveScheme(self.network)
scheme.append([NearestNeighborRepExStrategy(),
OneWayShootingStrategy(),
strategy])
root = scheme.move_decision_tree()
nshoot = len(scheme.movers['shooting'])
nrepex = len(scheme.movers['repex'])
for mover in scheme.movers['shooting']:
assert_almost_equal(scheme.choice_probability[mover], old_div(1.0,8.0))
for mover in scheme.movers['repex']:
assert_almost_equal(scheme.choice_probability[mover], old_div(1.0,16.0))
strategy.group_weights['shooting'] = 2.0
root = scheme.move_decision_tree(rebuild=True)
assert strategy.group_weights == {'shooting' : 2.0, 'repex' : 0.5}
(group_weights, mover_weights) = strategy.get_weights(
scheme=scheme,
sorted_movers=scheme.movers,
sort_weights_override=strategy.group_weights
)
assert group_weights == {'shooting' : 2.0, 'repex' : 0.5}
# everything within the group should have the same mover_weight
for group in scheme.movers:
group_sigs = [s for s in mover_weights if s[0]==group]
for sig in group_sigs:
assert mover_weights[sig] == mover_weights[group_sigs[0]]
choice_prob = strategy.choice_probability(
scheme, group_weights, mover_weights
)
for mover in scheme.movers['shooting']:
assert_almost_equal(choice_prob[mover], old_div(1.0,7.0))
for mover in scheme.movers['repex']:
assert_almost_equal(choice_prob[mover], old_div(1.0,28.0))
def test_get_weights_mover_weights_set(self):
strategy = OrganizeByMoveGroupStrategy()
scheme = MoveScheme(self.network)
scheme.append([NearestNeighborRepExStrategy(),
OneWayShootingStrategy(),
strategy])
# build it so there's an old version
strategy.group_weights['repex'] = 3.0
root = scheme.move_decision_tree()
(group_weights, mover_weights) = strategy.get_weights(
scheme=scheme,
sorted_movers=scheme.movers,
sort_weights_override=strategy.group_weights
)
assert group_weights == {'shooting' : 1.0, 'repex' : 3.0}