-
Notifications
You must be signed in to change notification settings - Fork 0
/
sampling.py
1032 lines (802 loc) · 35.5 KB
/
sampling.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
#!/usr/bin/env python3
# Copyright 2023 L3Harris Technologies, Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
import uuid
import json
import numpy as np
import networkx as nx
import matplotlib.pyplot as plt
def convert_index_to_coordinates(i: int, dimensions: np.ndarray) -> np.ndarray:
"""A utility function for converting square-grid lattice node
indexes to (x,y,z) positions.
This is only used for the square-grid lattice.
Args:
i (int): The node index
dimensions (np.ndarray): The dimensions of the lattice
(e.g., 4x4x4 is [4,4,4])
Returns:
np.ndarray: The [x,y,z] position.
"""
remainder = i
coords = np.zeros(len(dimensions))
for d in range(len(dimensions)):
coords[d] = int(remainder // np.prod(dimensions[d+1:]))
remainder = int(remainder % np.prod(dimensions[d+1:]))
return coords
def convert_coordinates_to_index(
coords: np.ndarray,
dimensions: np.ndarray
) -> int:
"""A utility function for converting square-grid lattice node
(x,y,z) positions to node indexes.
This is only used for the square-grid lattice.
Args:
coords (np.ndarray): The [x,y,z] position of the node.
dimensions (np.ndarray): The dimensions of the lattice
(e.g., 4x4x4 is [4,4,4])
Returns:
int: The index of the node.
"""
i = 0
for d in range(len(coords)):
i += coords[d]*np.prod(dimensions[d+1:])
return i
def convert_spin_config_to_state_index(spin_config: np.ndarray) -> int:
"""A utility function to convert a spin configuration to a state index.
Args:
spin_config (np.ndarray): An array of +1/-1 for each node.
Returns:
int: An index for the corresponding state.
"""
state = 0x0
for i in range(len(spin_config)):
if spin_config[i] == 1:
state = state | (0x1 << i)
return state
def convert_state_index_to_spin_config(n: int, state_index: int) -> np.ndarray:
"""A utility function to convert a state index to a spin configuration.
Args:
n (int): The number of nodes in the graph.
state_index (int): An index for the state.
Returns:
np.ndarray: A corresponding array of +1/-1 for each node.
"""
spin_config = np.zeros(n)
for i in range(n):
spin_config[i] = -1.0 + 2*bool(state_index & (0x1<<i))
return spin_config
def generate_random_spin_config(n: int) -> np.ndarray:
"""A utility function to generate a random array of [+1/-1] of length n.
Args:
n (int): The length of the array to be returned. Typically
corresponds to the number of
nodes in the graph.
Returns:
np.ndarray: An array of [+1/-1] of length n.
"""
return 2*(np.random.random(n) > 0.5*np.ones(n)) - np.ones(n)
def load_graph_instance_from_file(filename: str=None):
"""Loads the parameters of a graph model from a file.
Args:
filename (str, optional): The filename. Defaults to None.
Returns:
Graph: a sampling.Graph object.
"""
with open(filename,"rb") as f:
read_bytes = f.read()
f.close()
# convert the byte array to a dict
instance = json.loads(read_bytes)
G_nx = nx.node_link_graph(instance["graph_data"])
G = Graph(G_nx)
G.name = instance["metadata"]["graph_name"]
G.uuid = instance["metadata"]["graph_uuid"]
G.temperature_T = instance["metadata"]["temperature_T"]
G.k_B = instance["metadata"]["k_B"]
G.synchronize_attributes()
return G
def init_square_grid_nodes(dimensions: list=None) -> dict:
"""Creates a set of nodes for square-grid lattice graphs.
Args:
dimensions (list, optional): The dimensions of the lattice (e.g.,
4x4x4 is [4,4,4]). Defaults to None.
Returns:
dict: A dictionary of the nodes and related parameters of the graph.
"""
nodes = {}
n = np.prod(dimensions)
for i in range(n):
nodes[i] = {
"position":list((convert_index_to_coordinates(i, dimensions))),
"spin":1.0, # default is all +1.
"B":0.0 # default is no external field
}
return nodes
def init_square_grid_internal_edges(
nodes: dict=None,
distance_threshold: float=1.01
) -> dict:
"""Creates a set of edges for the nodes in a square-grid lattice.
Args:
nodes (dict, optional): A dictionary of nodes (index and position).
Typically provided by the `init_square_grid_nodes` function.
Defaults to None.
distance_threshold (float, optional): If nodes are less than this
distance apart, an edge is created between them.. Defaults to 1.01.
Returns:
dict: A set of edges and related parameters.
"""
edges = {}
for i in nodes:
for j in nodes:
if i < j:
xyz_i = np.array(nodes[i]["position"])
xyz_j = np.array(nodes[j]["position"])
distance = np.linalg.norm(xyz_i - xyz_j)
if distance < distance_threshold:
edges[(i,j)] = {
"weight":1.0, # default is 1.0
"periodic_boundary_edge":False # internal edge
}
return edges
def add_square_grid_periodic_boundary_edges(
dimensions: list=None,
nodes: dict=None,
edges: dict=None
) -> dict:
"""Creates additional periodc boundary edges for the square-grid lattice.
Args:
dimensions (list, optional): The dimensions of the lattice (e.g.,
4x4x4 is [4,4,4]). Defaults to None.
nodes (dict, optional): A dictionary of nodes (index and position).
Typically provided by the `init_square_grid_nodes` function.
Defaults to None.
edges (dict, optional): A dictionary of edges and related parameters.
Typically provided by the `init_square_grid_internal_edges` function.
Defaults to None.
Returns:
dict: A dictionary of edges and related parameters that includes the
original internal edges and the newly added periodic boundary edges.
"""
for i in nodes:
i_x = nodes[i]["position"][0]
i_y = nodes[i]["position"][1]
i_z = nodes[i]["position"][2]
if i_x == 0:
for j in nodes:
if i < j:
if nodes[j]["position"][0] == dimensions[0] - 1:
if nodes[j]["position"][1] == i_y:
if nodes[j]["position"][2] == i_z:
if (i, j) not in edges:
edges[(i,j)]={
"weight":1.0, # init to 1
"periodic_boundary_edge":True
}
break
if i_x == dimensions[0] - 1:
for j in nodes:
if i < j:
if nodes[j]["position"][0] == 0:
if nodes[j]["position"][1] == i_y:
if nodes[j]["position"][2] == i_z:
if (i, j) not in edges:
edges[(i,j)]={
"weight":1.0, # init to 1
"periodic_boundary_edge":True
}
break
if i_y == 0:
for j in nodes:
if i < j:
if nodes[j]["position"][0] == i_x:
if nodes[j]["position"][1] == dimensions[1] - 1:
if nodes[j]["position"][2] == i_z:
if (i, j) not in edges:
edges[(i,j)]={
"weight":1.0, # init to 1
"periodic_boundary_edge":True
}
break
if i_y == dimensions[1] - 1:
for j in nodes:
if i < j:
if nodes[j]["position"][0] == i_x:
if nodes[j]["position"][1] == 0:
if nodes[j]["position"][2] == i_z:
if (i, j) not in edges:
edges[(i,j)]={
"weight":1.0, # init to 1
"periodic_boundary_edge":True
}
break
if i_z == 0:
for j in nodes:
if i < j:
if nodes[j]["position"][0] == i_x:
if nodes[j]["position"][1] == i_y:
if nodes[j]["position"][2] == i_z:
if (i, j) not in edges:
edges[(i,j)]={
"weight":1.0, # init to 1
"periodic_boundary_edge":True
}
break
if i_z == dimensions[2] - 1:
for j in nodes:
if i < j:
if nodes[j]["position"][0] == i_x:
if nodes[j]["position"][1] == i_y:
if nodes[j]["position"][2] == i_z:
if (i, j) not in edges:
edges[(i,j)]={
"weight":1.0, # init to 1
"periodic_boundary_edge":True
}
break
return edges
class Graph(nx.Graph):
"""The Graph() object class extendes the NetworkX Graph
object class to include
additional fields, parameters and methods.
Args:
nx (_type_): extends the NetworkX Graph class.
"""
def __init__(
self,
nx_graph: nx.Graph=None,
nodes: dict=None,
edges: dict=None,
name: str="default_name",
user_defined_uuid: str=None
):
"""Initialize a sampling.Graph() object with various inputs.
If no input is given, an empty sampling.Graph() object is created.
Args:
nx_graph (nx.Graph, optional): Option 1: convert an existing
NetworkX graph. Defaults to None.
nodes (dict, optional): Option 2a: provide a set of nodes
(and edges). Defaults to None.
edges (dict, optional): Option 2b: provide a set of edges
(and nodes). Defaults to None.
name (str, optional): A user-defined name. Defaults to
"default_name".
uuid (str, optional): A user-defined UUID. If not specified,
one will be generated. Defaults to None.
"""
# initialize an empty NetworkX graph:
super(Graph, self).__init__(nx_graph)
self.name = name
self.temperature_T = None
self.k_B = None
if user_defined_uuid is None:
self.uuid = str(uuid.uuid4())
else:
self.uuid = user_defined_uuid
# after other parameters are updated,
# the following may be updated by calling the .synchronize_attributes() method.
self._has_external_field = None
self._consistent_external_field = None
self._edge_type = None # antiferromagnetic, ferromagnetic, spin-glass
self._consistent_edge_weights = None
self._beta = None
self._has_periodic_boundary_edges = None
self._max_degree = None
self._min_edge_weight_magnitude = None
self._max_edge_weight_magnitude = None
self._effective_min_edge_weight_magnitude = None # factors in beta
self._effective_max_edge_weight_magnitude = None # factors in beta
self._average_node_connectivity = None
self._is_planar = None
self._density = None
# _exact_probability_distribution is updated by calling the
# .brute_force_probability_distribution_calculation() method.
self._exact_probability_distribution = None
if nodes is not None:
self.load_nodes(nodes)
if edges is not None:
self.load_edges(edges)
def load_nodes(self, nodes: dict=None):
"""Load (merge) a set of nodes into a sampling.Graph() object.
Args:
nodes (dict, optional): A set of nodes
(index, position, spin, external_field B). Defaults to None.
"""
for i in nodes:
self.add_node(
i,
position=nodes[i]["position"],
spin=nodes[i]["spin"],
B=nodes[i]["B"]
)
self.synchronize_attributes()
def load_edges(self, edges: dict=None):
"""Load (merge) a set of edges into a sampling.Graph() object.
Args:
edges (dict, optional): A set of edges ((i,j), weight,
periodic_boundary_edge=(true/false)). Defaults to None.
"""
for e in edges:
i = e[0]
j = e[1]
self.add_edge(
i,
j,
weight=edges[e]["weight"],
periodic_boundary_edge=edges[e]["periodic_boundary_edge"]
)
self.synchronize_attributes()
def write_instance_to_file(self,
relative_path: str=None,
instance_uuid: str=None,
filename: str=None,
):
"""Write a sampling.Graph() instance to a file.
Args:
relative_path (str, optional): The path of the directory that the
instance will be written to. If the user does not specify, the
instance will be written to the current working directory.
Defaults to None.
instance_uuid (str, optional): A user-provided UUID.
If the user does not specify, it is generated. Defaults to None.
filename (str, optional): A user-provided filename.
If the user does not specify, it is automatically
generated based on the UUID. Defaults to None.
"""
self.synchronize_attributes()
if instance_uuid is None:
instance_uuid = str(uuid.uuid4())
if relative_path is None:
relative_path = "."
# we made the conscious decision to preserve and use the data structure
# provided/read by the NetworkX.node_link_data() function.
# other attributes appear in the metadata dictionary.
# specifically exporting a graph using the .node_link_data() function
# does NOT include .temperature_T and ._beta and some other attributes.
graph_data = nx.node_link_data(self)
benchmark_requirements = {
"num_samples_k":1000, # TODO: fixed at this time.
"time_limit_seconds":600 # TODO: fixed at this time.
}
metadata = self.get_summary_details()
metadata["instance_uuid"] = instance_uuid
aggregated_dictionary = {
"benchmark_requirements":benchmark_requirements,
"metadata":metadata,
"graph_data":graph_data
}
# write the dict as JSON to a file
if filename is None:
full_filename = relative_path + "/instance." + instance_uuid + ".json"
else:
full_filename = relative_path + "/" + filename
output_bytes = json.dumps(aggregated_dictionary).encode("utf8")
with open(full_filename,"wb") as f:
f.write(output_bytes)
f.close()
def plot(self, show_node_lables: bool=False) -> None:
"""A Matplotlib figure is rendered of the graph.
Args:
show_node_lables (bool, optional): Writes node index labels
on each node. Defaults to False.
"""
node_xyz = np.array([self.nodes[i]["position"] for i in self.nodes()])
edge_xyz = np.array([(self.nodes[e[0]]["position"], self.nodes[e[1]]["position"]) for e in self.edges()])
fig = plt.figure()
ax = fig.add_subplot(111, projection="3d")
ax.set_xlabel("$x$ position")
ax.set_ylabel("$y$ position")
ax.set_zlabel("$z$ position")
ax.set_title("System of Interest")
colors = np.array([self.nodes[i]["spin"] for i in self.nodes()])
ax.scatter(
*node_xyz.T,
c=colors,
s=100,
ec="w",
vmin=-1.0,
vmax=1.0
)
for vizedge in edge_xyz:
ax.plot(*vizedge.T, color="tab:gray")
if show_node_lables:
for i in self.nodes():
pos_i = self.nodes[i]["position"]
ax.text(pos_i[0], pos_i[1], pos_i[2], str(i))
fig.tight_layout()
plt.show()
def get_adjacency_matrix_with_zero_diagonal_as_numpy_matrix(self) -> np.ndarray:
"""A utility function to return the edge weight/interaction
strength as a matrix.
Returns:
np.ndarray: An n-by-n matrix of the edge interaction weights.
(n is the number of nodes in the graph.)
"""
adj_matrix = nx.to_numpy_array(self)
external_field_B = np.array([ self.nodes[i]["B"] for i in range(len(self.nodes()))])
return adj_matrix, external_field_B
def brute_force_probability_distribution_calculation(self):
"""Calculates (or updates) Graph._exact_probability_distribution
in place.
This method should NOT be called if the number of nodes n is large.
"""
self.synchronize_attributes()
n = len(self.nodes())
N = 2**n
self._exact_probability_distribution = np.zeros(N)
for state_index in range(N):
spin_config = convert_state_index_to_spin_config(n,state_index)
self.set_spins(spin_config)
E_state = self.get_energy()
self._exact_probability_distribution[state_index] = np.exp(-1.0*self._beta*E_state)
Z = sum(self._exact_probability_distribution) # calculate partition function
self._exact_probability_distribution *= (1/Z) # normalize by partition function
def synchronize_attributes(self):
"""This method updates and synchronizes all
attributes of the sampling.Graph object
that the user should not directly modify
such as ._beta, ._edge_type.
This method is typically called before/after
a sampling.Graph object is written/read from a file.
This method should be called manually if a
user edits any attributes.
"""
if (self.temperature_T is not None) and (self.k_B is not None):
self._beta = 1/(self.temperature_T*self.k_B)
else:
self._beta = None
self._has_external_field = False
for i in self.nodes():
if np.abs(self.nodes[i]["B"]) > 1e-12:
self._has_external_field = True
break
if len(self.nodes()) > 0:
self._consistent_external_field = True
val = self.nodes[0]["B"]
for i in self.nodes():
if np.abs(self.nodes[i]["B"] - val) > 1e-12:
self._consistent_external_field = False
break
if len(self.edges()) > 0:
self._consistent_edge_weights = True
edge_1 = list(self.edges())[0]
for edge_2 in self.edges():
if edge_1 != edge_2:
w1 = self.edges[edge_1]["weight"]
w2 = self.edges[edge_2]["weight"]
if np.abs(w1 - w2) > 1e-12:
self._consistent_edge_weights = False
break
self._has_periodic_boundary_edges = False
for edge in self.edges():
if self.edges[edge]["periodic_boundary_edge"]:
self._has_periodic_boundary_edges = True
# determine ._edge_type \in {ferromagnetic, antiferromagnetic, spin-glass}
all_positive_weights = True
for edge in self.edges():
if self.edges[edge]["weight"] < -1e-12:
all_positive_weights = False
break
all_negative_weights = True
for edge in self.edges():
if self.edges[edge]["weight"] > 1e-12:
all_negative_weights = False
if all_positive_weights:
self._edge_type = "ferromagnetic"
if all_negative_weights:
self._edge_type = "antiferromagnetic"
if (not all_positive_weights) and (not all_negative_weights):
self._edge_type = "spin-glass"
# update other features of the graph
if len(self.edges()) > 0:
self._min_edge_weight_magnitude = np.min([np.abs(self.edges[e]["weight"]) for e in self.edges()])
if len(self.edges()) > 0:
self._max_edge_weight_magnitude = np.max([np.abs(self.edges[e]["weight"]) for e in self.edges()])
if self._beta is not None and self._min_edge_weight_magnitude is not None:
self._effective_min_edge_weight_magnitude = self._beta*self._min_edge_weight_magnitude
if self._beta is not None and self._max_edge_weight_magnitude is not None:
self._effective_max_edge_weight_magnitude = self._beta*self._max_edge_weight_magnitude
if len(self.nodes()) > 0:
self._max_degree = int(np.max([self.degree[i] for i in self.nodes()]))
self._average_node_connectivity = nx.average_node_connectivity(self)
self._density = nx.density(self)
self._is_planar = nx.is_planar(self)
def get_summary_details(self) -> dict:
"""Provides a dictionary of metadata
about a sampling.Graph object.
Returns:
dict: contains metadata and other
parameters of the sampling.Graph object.
"""
self.synchronize_attributes()
metadata = {}
metadata["graph_name"] = self.name
metadata["graph_uuid"] = self.uuid
metadata["number_of_nodes"] = len(self.nodes())
metadata["number_of_edges"] = len(self.edges())
metadata["temperature_T"] = self.temperature_T
metadata["k_B"] = self.k_B
metadata["beta"] = self._beta
metadata["edge_type"] = self._edge_type
metadata["consistent_edge_weights"] = self._consistent_edge_weights
if self._consistent_edge_weights:
e = list(self.edges())[0]
metadata["edge_weight"] = self.edges[e]["weight"]
else:
metadata["edge_weight"] = "varies"
metadata["has_periodic_boundary_edges"] = self._has_periodic_boundary_edges
metadata["has_nonzero_external_field"] = self._has_external_field
metadata["consistent_external_field"] = self._consistent_external_field
if self._consistent_external_field:
i = list(self.nodes())[0]
metadata["external_field"] = self.nodes[i]["B"]
else:
metadata["external_field"] = "varies"
metadata["max_degree"] = self._max_degree
metadata["min_edge_weight_magnitude"] = self._min_edge_weight_magnitude
metadata["max_edge_weight_magnitude"] = self._max_edge_weight_magnitude
metadata["effective_min_edge_weight_magnitude"] = self._effective_min_edge_weight_magnitude
metadata["effective_max_edge_weight_magnitude"] = self._effective_max_edge_weight_magnitude
metadata["average_node_connectivity"] = self._average_node_connectivity
metadata["is_planar"] = self._is_planar
metadata["density"] = self._density
return metadata
def set_spins(self, spin_config: np.ndarray):
"""Updates the spin configuration of a graph.
Args:
spin_config (np.ndarray): An array of [+1/-1] of
length n (where n is the number of nodes).
"""
for i in self.nodes():
self.nodes[i]["spin"] = spin_config[i]
def get_spins(self) -> np.ndarray:
"""Returns the current spin configuration.
Returns:
np.ndarray: An array of [+1/-1] of length n
(where n is the number of nodes) that describes
the current spin configuration.
"""
return np.array([self.nodes[i]["spin"] for i in self.nodes()])
def set_external_field(self, B: np.ndarray):
"""Sets the value of the external field `B` property of nodes.
Args:
B (np.ndarray): An array of length n (where n is the number of nodes)
that describes the strength of the external field applied to each node.
Values may be positive or negative.
"""
for i in self.nodes():
self.nodes[i]["B"] = B[i]
def get_external_field(self) -> np.ndarray:
"""Returns the current values of the external field
Returns:
np.ndarray: An array of length n (where n is the number of nodes).
"""
return np.array([self.nodes[i]["B"] for i in self.nodes()])
def set_couplings(self, J: dict):
"""Sets the values of the coupling strength `weight` parameter of edges.
Args:
J (dict): A dictionary with keys of the form (i,j) and values describing the weight to set for edge (i,j)
"""
if len(self.edges()) == 0:
raise Exception("Error: no edges in the graph!")
for e in self.edges():
i = e[0]
j = e[1]
self.edges[e]["weight"] = J[i, j]
def get_couplings(self) -> dict:
"""Returns the current values for coupling strength `weight` parameter of edges.
Returns:
dict: A dictionary with keys of the form (i,j) and values describing the weight to set for edge (i,j)
"""
return {e:self.edges[e]["weight"] for e in self.edges()}
def get_energy(self) -> float :
"""Returns the current energy of the graph based on the current spin configuration.
$H(\sigma) = - \sum_{i~j} J_{ij} \sigma_i \sigma_j - \sum_i B_i \sigma_i$
Note that we are going with the classic definition with the minus signs.
Note that we allow spin for each node to be +1, -1, but in our "constructing" starting spin_configs we allow spin to be 0 for unset nodes.
J_ij > 0 implies ferromagnetic: neighbors prefer to be the same spin.
J_ij < 0 implies antiferromagnetic: neighbors prefer to be opposite spins.
Returns:
float: The energy of the system as a real number.
"""
energy_sum = 0.0
for i in self.nodes():
energy_sum -= self.nodes[i]["B"]*self.nodes[i]["spin"]
for e in self.edges():
i = e[0]
j = e[1]
energy_sum -= self.edges[(i, j)]["weight"] * \
self.nodes[i]["spin"]*self.nodes[j]["spin"]
return energy_sum
def generate_mcmc_sample_v1(self,
num_steps: int,
rng_seed: int=None,
print_status: bool=None
) -> tuple:
"""This will generate a MCMC sample from the graph.
The Markov chain will start at whatever state the current spin config is.
In addition to returning the energy and spin configuration of the sample,
the spin configuration of the graph will be updated in place and can be
read later.
Args:
num_steps (int): The integer number of steps to take on the chain.
rng_seed (int, optional): The random number seed to use. This may
be set for repeatability or for Markov chain coupling. Defaults to
None.
print_status (bool, optional): Prints progress to stdout. Defaults
to None.
Raises:
Exception: WARNING! Setting RNG seed is not implemented and will
throw and error.
Returns:
tuple: (final energy as a float, final spin configuration
as an np.ndarray)
"""
if rng_seed is not None:
raise Exception("Error: fixed RNG seed not implemented!")
n = len(self.nodes())
i = 0
for step in range(num_steps):
self.generate_mcmc_mh_step(
target_spin=i
)
i += 1
i = i % n
if print_status:
if step % 50 == 0:
print("step:",step,"/",num_steps)
return self.get_energy(), self.get_spins()
def generate_mcmc_mh_step(self,
target_spin: int=None,
uniform_random_number: float=None,
) -> tuple :
"""Executes a single step in a Monte Carlo Markov Chain (MCMC)
sampling procedure.
Args:
target_spin (int, optional): The node index i to propose
flipping the spin for. Defaults to None.
uniform_random_number (float, optional): A uniform random number
used to accept/reject the proposed node spin flip. Defaults to None.
Returns:
tuple: (energy after the step as a float,
spin configuration after the step as an np.ndarray)
"""
if uniform_random_number is None:
uniform_random_number = np.random.random()
current_energy = self.get_energy()
self.nodes[target_spin]["spin"] *= -1 # flip sign
proposed_energy = self.get_energy()
if proposed_energy < (current_energy - 1e-9):
# accept proposed state because it is lower energy.
current_energy = proposed_energy
else:
threshold = np.exp(-1.0*self._beta*(proposed_energy - current_energy))
if uniform_random_number < threshold:
# accept proposed state even though it is higher energy
current_energy = proposed_energy
else:
# do NOT accept proposed state
# keep E_current set as is (no update)
self.nodes[target_spin]["spin"] *= -1 # revert to keep the "current" state
return current_energy, self.get_spins()
################################################################################
# Functions used to generate greedy randomized spin configurations
def is_complete_spin_config(spin_config: np.ndarray) -> bool:
"""A utility function to check that a spin configuration is completely specified.
Used as a subroutine to `generate_randomized_greedy_spin_config`.
Args:
spin_config (np.ndarray): An array of [+1/-1] of length n.
Returns:
bool: True if the spin configuration has +1/-1 for each node.
"""
n = len(spin_config)
return bool(np.prod(abs(spin_config) > 1e-6*np.ones(n)))
def generate_randomized_greedy_spin_config(
G: Graph=None,
plot_progress: bool=False
) -> np.ndarray:
"""A routine to generate spin configurations that prefer low energy
configurations.
Args:
G (sampling.Graph, optional): A samping.Graph() object. Defaults to None.
plot_progress (bool, optional): Print progress to stdout. Defaults to False.
Returns:
np.ndarray: An array of [+1/-1] of length n representing the
constructed spin configuration.
"""
n = len(G.nodes())
spin_config = np.zeros(n) # an empty config
i = np.random.randint(0,n-1) # randomly pick first node i
spin_config[i] = 2*(np.random.random() > 0.5) - 1.0 # randomly set node i to +1.0 or -1.0.
G.set_spins(spin_config)
if plot_progress:
G.plot()
print("spin config:",G.get_spins())
print("Energy:",G.get_energy())
while not is_complete_spin_config(spin_config):
nbors, conn_nbors = collect_set_of_fixed_spin_neighbors(
G=G,
spin_config=spin_config
)
#randomly pick a candidate neighbor
# TODO: weighted random pick by conn_nbors
if len(nbors) == 1:
j = nbors[0]
else:
nbor_index = np.random.randint(0,len(nbors)-1)
j = nbors[nbor_index]
candidate_spin_config_down = spin_config.copy()
candidate_spin_config_down[j] = -1.0
candidate_spin_config_up = spin_config.copy()
candidate_spin_config_up[j] = 1.0
G.set_spins(candidate_spin_config_down)
E_candidate_down = G.get_energy()
e_down = np.exp(-1.0*G._beta*E_candidate_down)
G.set_spins(candidate_spin_config_up)
E_candidate_up = G.get_energy()
e_up = np.exp(-1.0*G._beta*E_candidate_up)
prob_down = e_down / (e_down + e_up)
u = np.random.rand()
if u < prob_down:
spin_config[j] = -1.0
else:
spin_config[j] = 1.0
G.set_spins(spin_config)
if plot_progress:
G.plot()
print("spin config:",G.get_spins())
print("Energy:",G.get_energy())
return spin_config
def collect_set_of_fixed_spin_neighbors(
G: Graph=None,
spin_config: np.ndarray=None
) -> tuple:
"""A utility function to identify neighbors of a node.
Used as a subroutine to `generate_randomized_greedy_spin_config`.
Args:
G (sampling.Graph, optional): A samping.Graph() object. Defaults to
None.
spin_config (np.ndarray, optional): An array of [+1/-1] of length n
representing the
constructed spin configuration. Defaults to None.
Returns:
tuple: ( np.ndarray of neighbors (integers) , np.ndarray of how many
connections each neighbor has )
"""
n = len(G.nodes())
set_of_neighbors_with_no_spin_yet = []