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
|
"""!
@brief Templates for tests of SyncPR (oscillatory network based on Kuramoto model for pattern recognition).
@authors Andrei Novikov (pyclustering@yandex.ru)
@date 2014-2020
@copyright BSD-3-Clause
"""
# Generate images without having a window appear.
import matplotlib;
matplotlib.use('Agg');
from pyclustering.nnet import solve_type;
from pyclustering.nnet.syncpr import syncpr, syncpr_visualizer;
class SyncprTestTemplates:
@staticmethod
def templateOutputDynamic(solver, ccore):
net = syncpr(5, 0.1, 0.1, ccore);
output_dynamic = net.simulate(10, 10, [-1, 1, -1, 1, -1], solver, True);
assert len(output_dynamic) == 11; # 10 steps without initial values.
@staticmethod
def templateOutputDynamicLengthStaticSimulation(collect_flag, ccore_flag):
net = syncpr(5, 0.1, 0.1, ccore_flag);
output_dynamic = net.simulate_static(10, 10, [-1, 1, -1, 1, -1], solution = solve_type.FAST, collect_dynamic = collect_flag);
if (collect_flag is True):
assert len(output_dynamic) == 11; # 10 steps without initial values.
else:
assert len(output_dynamic) == 1;
@staticmethod
def templateOutputDynamicLengthDynamicSimulation(collect_flag, ccore_flag):
net = syncpr(5, 0.1, 0.1, ccore_flag);
output_dynamic = net.simulate_dynamic([-1, 1, -1, 1, -1], solution = solve_type.FAST, collect_dynamic = collect_flag);
if (collect_flag is True):
assert len(output_dynamic) > 1;
else:
assert len(output_dynamic) == 1;
@staticmethod
def templateIncorrectPatternForSimulation(pattern, ccore_flag):
net = syncpr(10, 0.1, 0.1, ccore=ccore_flag);
try: net.simulate(10, 10, pattern);
except: return;
assert False;
@staticmethod
def templateTrainNetworkAndRecognizePattern(ccore_flag):
net = syncpr(10, 0.1, 0.1, ccore_flag);
patterns = [];
patterns += [ [1, 1, 1, 1, 1, -1, -1, -1, -1, -1] ];
patterns += [ [-1, -1, -1, -1, -1, 1, 1, 1, 1, 1] ];
net.train(patterns);
# recognize it
for i in range(len(patterns)):
output_dynamic = net.simulate(10, 10, patterns[i], solve_type.RK4, True);
ensembles = output_dynamic.allocate_sync_ensembles(0.5);
assert len(ensembles) == 2;
assert len(ensembles[0]) == len(ensembles[1]);
# sort results
ensembles[0].sort();
ensembles[1].sort();
assert (ensembles[0] == [0, 1, 2, 3, 4]) or (ensembles[0] == [5, 6, 7, 8, 9]);
assert (ensembles[1] == [0, 1, 2, 3, 4]) or (ensembles[1] == [5, 6, 7, 8, 9]);
@staticmethod
def templateIncorrectPatternForTraining(patterns, ccore_flag):
net = syncpr(10, 0.1, 0.1, ccore_flag);
try: net.train(patterns);
except: return;
assert False;
@staticmethod
def templatePatternVisualizer(collect_dynamic, ccore_flag = False):
net = syncpr(5, 0.1, 0.1, ccore = ccore_flag);
output_dynamic = net.simulate(10, 10, [-1, 1, -1, 1, -1], solve_type.RK4, collect_dynamic);
syncpr_visualizer.show_pattern(output_dynamic, 5, 2);
syncpr_visualizer.animate_pattern_recognition(output_dynamic, 1, 5);
@staticmethod
def templateMemoryOrder(ccore_flag):
net = syncpr(10, 0.1, 0.1, ccore_flag);
patterns = [];
patterns += [ [1, 1, 1, 1, 1, -1, -1, -1, -1, -1] ];
patterns += [ [-1, -1, -1, -1, -1, 1, 1, 1, 1, 1] ];
net.train(patterns);
assert net.memory_order(patterns[0]) < 0.8;
assert net.memory_order(patterns[1]) < 0.8;
for pattern in patterns:
net.simulate(20, 10, pattern, solve_type.RK4);
memory_order = net.memory_order(pattern);
assert (memory_order > 0.95) and (memory_order <= 1.000005);
@staticmethod
def templateStaticSimulation(ccore_falg):
net = syncpr(10, 0.1, 0.1, ccore_falg);
patterns = [];
patterns += [ [1, 1, 1, 1, 1, -1, -1, -1, -1, -1] ];
patterns += [ [-1, -1, -1, -1, -1, 1, 1, 1, 1, 1] ];
net.train(patterns);
net.simulate_static(20, 10, patterns[0], solve_type.RK4);
memory_order = net.memory_order(patterns[0]);
assert (memory_order > 0.95) and (memory_order <= 1.000005);
@staticmethod
def templateDynamicSimulation(ccore_flag):
net = syncpr(10, 0.1, 0.1, ccore_flag);
patterns = [];
patterns += [ [1, 1, 1, 1, 1, -1, -1, -1, -1, -1] ];
patterns += [ [-1, -1, -1, -1, -1, 1, 1, 1, 1, 1] ];
net.train(patterns);
net.simulate_dynamic(patterns[0], order = 0.998, solution = solve_type.RK4);
memory_order = net.memory_order(patterns[0]);
assert (memory_order > 0.998) and (memory_order <= 1.0);
@staticmethod
def templateGlobalSyncOrder(ccore_flag):
net = syncpr(10, 0.1, 0.1, ccore_flag);
patterns = [ [1, 1, 1, 1, 1, -1, -1, -1, -1, -1] ];
patterns += [ [-1, -1, -1, -1, -1, 1, 1, 1, 1, 1] ];
global_sync_order = net.sync_order();
assert (global_sync_order < 1.0) and (global_sync_order > 0.0);
net.train(patterns);
global_sync_order = net.sync_order();
assert (global_sync_order < 1.0) and (global_sync_order > 0.0);
@staticmethod
def templateLocalSyncOrder(ccore_flag):
net = syncpr(10, 0.1, 0.1, ccore_flag);
patterns = [ [1, 1, 1, 1, 1, -1, -1, -1, -1, -1] ];
patterns += [ [-1, -1, -1, -1, -1, 1, 1, 1, 1, 1] ];
local_sync_order = net.sync_local_order();
assert (local_sync_order < 1.0) and (local_sync_order > 0.0);
net.train(patterns);
local_sync_order = net.sync_local_order();
assert (local_sync_order < 1.0) and (local_sync_order > 0.0);
@staticmethod
def templateIncorrectPatternValues(ccore_flag):
patterns = [];
patterns += [ [2, 1, 1, 1, 1, -1, -1, -1, -1, -1] ];
patterns += [ [-1, -2, -1, -1, -1, 1, 1, 1, 1, 1] ];
SyncprTestTemplates.templateIncorrectPatternForTraining(patterns, ccore_flag);
|
