|
1
|
|
|
"""!
|
|
2
|
|
|
|
|
3
|
|
|
@brief Neural Network: Local Excitatory Global Inhibitory Oscillatory Network (LEGION)
|
|
4
|
|
|
@details Based on article description:
|
|
5
|
|
|
- D.Wang, D.Terman. Image Segmentation Based on Oscillatory Correlation. 1997.
|
|
6
|
|
|
- D.Wang, D.Terman. Locally Excitatory Globally Inhibitory Oscillator Networks. 1995.
|
|
7
|
|
|
|
|
8
|
|
|
@authors Andrei Novikov ([email protected])
|
|
9
|
|
|
@date 2014-2017
|
|
10
|
|
|
@copyright GNU Public License
|
|
11
|
|
|
|
|
12
|
|
|
@cond GNU_PUBLIC_LICENSE
|
|
13
|
|
|
PyClustering is free software: you can redistribute it and/or modify
|
|
14
|
|
|
it under the terms of the GNU General Public License as published by
|
|
15
|
|
|
the Free Software Foundation, either version 3 of the License, or
|
|
16
|
|
|
(at your option) any later version.
|
|
17
|
|
|
|
|
18
|
|
|
PyClustering is distributed in the hope that it will be useful,
|
|
19
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
20
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
21
|
|
|
GNU General Public License for more details.
|
|
22
|
|
|
|
|
23
|
|
|
You should have received a copy of the GNU General Public License
|
|
24
|
|
|
along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
25
|
|
|
@endcond
|
|
26
|
|
|
|
|
27
|
|
|
"""
|
|
28
|
|
|
|
|
29
|
|
|
import numpy;
|
|
|
|
|
|
|
30
|
|
|
import math;
|
|
31
|
|
|
import random;
|
|
32
|
|
|
|
|
33
|
|
|
import pyclustering.core.legion_wrapper as wrapper;
|
|
34
|
|
|
|
|
35
|
|
|
from pyclustering.nnet import *;
|
|
|
|
|
|
|
36
|
|
|
|
|
37
|
|
|
from pyclustering.utils import heaviside, allocate_sync_ensembles;
|
|
38
|
|
|
|
|
39
|
|
|
from scipy.integrate import odeint;
|
|
|
|
|
|
|
40
|
|
|
|
|
41
|
|
|
|
|
42
|
|
|
class legion_parameters:
|
|
43
|
|
|
"""!
|
|
44
|
|
|
@brief Describes parameters of LEGION.
|
|
45
|
|
|
@details Contained parameters affect on output dynamic of each oscillator of the network.
|
|
46
|
|
|
|
|
47
|
|
|
@see legion_network
|
|
48
|
|
|
|
|
49
|
|
|
"""
|
|
50
|
|
|
|
|
51
|
|
|
def __init__(self):
|
|
52
|
|
|
"""!
|
|
53
|
|
|
@brief Default constructor of parameters for LEGION (local excitatory global inhibitory oscillatory network).
|
|
54
|
|
|
@details Constructor initializes parameters by default non-zero values that can be
|
|
55
|
|
|
used for simple simulation.
|
|
56
|
|
|
"""
|
|
57
|
|
|
|
|
58
|
|
|
## Coefficient that affects intrinsic inhibitor of each oscillator. Should be the same as 'alpha'.
|
|
59
|
|
|
self.eps = 0.02;
|
|
60
|
|
|
|
|
61
|
|
|
## Coefficient is chosen to be on the same order of magnitude as 'eps'. Affects on exponential function that decays on a slow time scale.
|
|
62
|
|
|
self.alpha = 0.005;
|
|
63
|
|
|
|
|
64
|
|
|
## Coefficient that is used to control the ratio of the times that the solution spends in these two phases. For a larger value of g, the solution spends a shorter time in the active phase.
|
|
65
|
|
|
self.gamma = 6.0;
|
|
66
|
|
|
|
|
67
|
|
|
## Coefficient that affects on intrinsic inhibitor of each oscillator. Specifies the steepness of the sigmoid function.
|
|
68
|
|
|
self.betta = 0.1;
|
|
69
|
|
|
|
|
70
|
|
|
## Scale coefficient that is used by potential, should be greater than 0.
|
|
71
|
|
|
self.lamda = 0.1;
|
|
72
|
|
|
|
|
73
|
|
|
## Threshold that should be exceeded by a potential to switch on potential.
|
|
74
|
|
|
self.teta = 0.9;
|
|
75
|
|
|
|
|
76
|
|
|
## Threshold that should be exceeded by a single oscillator to affect its neighbors.
|
|
77
|
|
|
self.teta_x = -1.5;
|
|
78
|
|
|
|
|
79
|
|
|
## Threshold that should be exceeded to activate potential. If potential less than the threshold then potential is relaxed to 0 on time scale 'mu'.
|
|
80
|
|
|
self.teta_p = 1.5;
|
|
81
|
|
|
|
|
82
|
|
|
## Threshold that should be exceeded by any oscillator to activate global inhibitor.
|
|
83
|
|
|
self.teta_xz = 0.1;
|
|
84
|
|
|
|
|
85
|
|
|
## Threshold that should be exceeded to affect on a oscillator by the global inhibitor.
|
|
86
|
|
|
self.teta_zx = 0.1;
|
|
87
|
|
|
|
|
88
|
|
|
## Weight of permanent connections.
|
|
89
|
|
|
self.T = 2.0;
|
|
90
|
|
|
|
|
91
|
|
|
## Defines time scaling of relaxing of oscillator potential.
|
|
92
|
|
|
self.mu = 0.01;
|
|
93
|
|
|
|
|
94
|
|
|
## Weight of global inhibitory connections.
|
|
95
|
|
|
self.Wz = 1.5;
|
|
96
|
|
|
|
|
97
|
|
|
## Total dynamic weights to a single oscillator from neighbors. Sum of weights of dynamic connections to a single oscillator can not be bigger than Wt.
|
|
98
|
|
|
self.Wt = 8.0;
|
|
99
|
|
|
|
|
100
|
|
|
## Rate at which the global inhibitor reacts to the stimulation from the oscillator network.
|
|
101
|
|
|
self.fi = 3.0;
|
|
102
|
|
|
|
|
103
|
|
|
## Multiplier of oscillator noise. Plays important role in desynchronization process.
|
|
104
|
|
|
self.ro = 0.02;
|
|
105
|
|
|
|
|
106
|
|
|
## Value of external stimulus.
|
|
107
|
|
|
self.I = 0.2;
|
|
108
|
|
|
|
|
109
|
|
|
## Defines whether to use potentional of oscillator or not.
|
|
110
|
|
|
self.ENABLE_POTENTIONAL = True;
|
|
111
|
|
|
|
|
112
|
|
|
|
|
113
|
|
|
class legion_dynamic:
|
|
114
|
|
|
"""!
|
|
115
|
|
|
@brief Represents output dynamic of LEGION.
|
|
116
|
|
|
|
|
117
|
|
|
"""
|
|
118
|
|
|
|
|
119
|
|
|
@property
|
|
120
|
|
|
def output(self):
|
|
121
|
|
|
"""!
|
|
122
|
|
|
@brief Returns output dynamic of the network.
|
|
123
|
|
|
|
|
124
|
|
|
"""
|
|
125
|
|
|
if (self.__ccore_legion_dynamic_pointer is not None):
|
|
126
|
|
|
return wrapper.legion_dynamic_get_output(self.__ccore_legion_dynamic_pointer);
|
|
127
|
|
|
|
|
128
|
|
|
return self.__output;
|
|
129
|
|
|
|
|
130
|
|
|
|
|
131
|
|
|
@property
|
|
132
|
|
|
def inhibitor(self):
|
|
133
|
|
|
"""!
|
|
134
|
|
|
@brief Returns output dynamic of the global inhibitor of the network.
|
|
135
|
|
|
|
|
136
|
|
|
"""
|
|
137
|
|
|
|
|
138
|
|
|
if (self.__ccore_legion_dynamic_pointer is not None):
|
|
139
|
|
|
return wrapper.legion_dynamic_get_inhibitory_output(self.__ccore_legion_dynamic_pointer);
|
|
140
|
|
|
|
|
141
|
|
|
return self.__inhibitor;
|
|
142
|
|
|
|
|
143
|
|
|
|
|
144
|
|
|
@property
|
|
145
|
|
|
def time(self):
|
|
146
|
|
|
"""!
|
|
147
|
|
|
@brief Returns simulation time.
|
|
148
|
|
|
|
|
149
|
|
|
"""
|
|
150
|
|
|
if (self.__ccore_legion_dynamic_pointer is not None):
|
|
151
|
|
|
return wrapper.legion_dynamic_get_time(self.__ccore_legion_dynamic_pointer);
|
|
152
|
|
|
|
|
153
|
|
|
return list(range(len(self)));
|
|
154
|
|
|
|
|
155
|
|
|
|
|
156
|
|
|
def __init__(self, output, inhibitor, time, ccore = None):
|
|
157
|
|
|
"""!
|
|
158
|
|
|
@brief Constructor of legion dynamic.
|
|
159
|
|
|
|
|
160
|
|
|
@param[in] output (list): Output dynamic of the network represented by excitatory values of oscillators.
|
|
161
|
|
|
@param[in] inhibitor (list): Output dynamic of the global inhibitor of the network.
|
|
162
|
|
|
@param[in] time (list): Simulation time.
|
|
163
|
|
|
@param[in] ccore (POINTER): Pointer to CCORE legion_dynamic. If it is specified then others arguments can be omitted.
|
|
164
|
|
|
|
|
165
|
|
|
"""
|
|
166
|
|
|
|
|
167
|
|
|
self.__output = output;
|
|
168
|
|
|
self.__inhibitor = inhibitor;
|
|
169
|
|
|
self._time = time;
|
|
170
|
|
|
|
|
171
|
|
|
self.__ccore_legion_dynamic_pointer = ccore;
|
|
172
|
|
|
|
|
173
|
|
|
|
|
174
|
|
|
def __del__(self):
|
|
175
|
|
|
"""!
|
|
176
|
|
|
@brief Destructor of the dynamic of the legion network.
|
|
177
|
|
|
|
|
178
|
|
|
"""
|
|
179
|
|
|
if (self.__ccore_legion_dynamic_pointer is not None):
|
|
180
|
|
|
wrapper.legion_dynamic_destroy(self.__ccore_legion_dynamic_pointer);
|
|
181
|
|
|
|
|
182
|
|
|
|
|
183
|
|
|
def __len__(self):
|
|
184
|
|
|
"""!
|
|
185
|
|
|
@brief Returns length of output dynamic.
|
|
186
|
|
|
|
|
187
|
|
|
"""
|
|
188
|
|
|
if (self.__ccore_legion_dynamic_pointer is not None):
|
|
189
|
|
|
return wrapper.legion_dynamic_get_size(self.__ccore_legion_dynamic_pointer);
|
|
190
|
|
|
|
|
191
|
|
|
return len(self._time);
|
|
192
|
|
|
|
|
193
|
|
|
|
|
194
|
|
|
def allocate_sync_ensembles(self, tolerance = 0.1):
|
|
195
|
|
|
"""!
|
|
196
|
|
|
@brief Allocate clusters in line with ensembles of synchronous oscillators where each synchronous ensemble corresponds to only one cluster.
|
|
197
|
|
|
|
|
198
|
|
|
@param[in] tolerance (double): Maximum error for allocation of synchronous ensemble oscillators.
|
|
199
|
|
|
|
|
200
|
|
|
@return (list) Grours of indexes of synchronous oscillators, for example, [ [index_osc1, index_osc3], [index_osc2], [index_osc4, index_osc5] ].
|
|
201
|
|
|
|
|
202
|
|
|
"""
|
|
203
|
|
|
|
|
204
|
|
|
if (self.__ccore_legion_dynamic_pointer is not None):
|
|
205
|
|
|
self.__output = wrapper.legion_dynamic_get_output(self.__ccore_legion_dynamic_pointer);
|
|
206
|
|
|
|
|
207
|
|
|
return allocate_sync_ensembles(self.__output, tolerance);
|
|
208
|
|
|
|
|
209
|
|
|
|
|
210
|
|
|
class legion_network(network):
|
|
211
|
|
|
"""!
|
|
212
|
|
|
@brief Local excitatory global inhibitory oscillatory network (LEGION) that uses relaxation oscillator
|
|
213
|
|
|
based on Van der Pol model.
|
|
214
|
|
|
|
|
215
|
|
|
@details The model uses global inhibitor to de-synchronize synchronous ensembles of oscillators.
|
|
216
|
|
|
|
|
217
|
|
|
Example:
|
|
218
|
|
|
@code
|
|
219
|
|
|
# Create parameters of the network
|
|
220
|
|
|
parameters = legion_parameters();
|
|
221
|
|
|
parameters.Wt = 4.0;
|
|
222
|
|
|
|
|
223
|
|
|
# Create stimulus
|
|
224
|
|
|
stimulus = [1, 1, 0, 0, 0, 1, 1, 1];
|
|
225
|
|
|
|
|
226
|
|
|
# Create the network (use CCORE for fast solving)
|
|
227
|
|
|
net = legion_network(len(stimulus), parameters, conn_type.GRID_FOUR, ccore = True);
|
|
228
|
|
|
|
|
229
|
|
|
# Simulate network - result of simulation is output dynamic of the network
|
|
230
|
|
|
output_dynamic = net.simulate(1000, 750, stimulus);
|
|
231
|
|
|
|
|
232
|
|
|
# Draw output dynamic
|
|
233
|
|
|
draw_dynamics(output_dynamic.time, output_dynamic.output, x_title = "Time", y_title = "x(t)");
|
|
234
|
|
|
@endcode
|
|
235
|
|
|
|
|
236
|
|
|
"""
|
|
237
|
|
|
|
|
238
|
|
|
def __init__(self, num_osc, parameters = None, type_conn = conn_type.ALL_TO_ALL, type_conn_represent = conn_represent.MATRIX, ccore = False):
|
|
239
|
|
|
"""!
|
|
240
|
|
|
@brief Constructor of oscillatory network LEGION (local excitatory global inhibitory oscillatory network).
|
|
241
|
|
|
|
|
242
|
|
|
@param[in] num_osc (uint): Number of oscillators in the network.
|
|
243
|
|
|
@param[in] parameters (legion_parameters): Parameters of the network that are defined by structure 'legion_parameters'.
|
|
244
|
|
|
@param[in] type_conn (conn_type): Type of connection between oscillators in the network.
|
|
245
|
|
|
@param[in] type_conn_represent (conn_represent): Internal representation of connection in the network: matrix or list.
|
|
246
|
|
|
@param[in] ccore (bool): If True then all interaction with object will be performed via CCORE library (C++ implementation of pyclustering).
|
|
247
|
|
|
|
|
248
|
|
|
"""
|
|
249
|
|
|
|
|
250
|
|
|
self._params = None; # parameters of the network
|
|
251
|
|
|
|
|
252
|
|
|
self.__ccore_legion_pointer = None;
|
|
253
|
|
|
self._params = parameters;
|
|
254
|
|
|
|
|
255
|
|
|
# set parameters of the network
|
|
256
|
|
|
if (self._params is None):
|
|
257
|
|
|
self._params = legion_parameters();
|
|
258
|
|
|
|
|
259
|
|
|
if (ccore is True):
|
|
260
|
|
|
self.__ccore_legion_pointer = wrapper.legion_create(num_osc, type_conn, self._params);
|
|
261
|
|
|
|
|
262
|
|
|
else:
|
|
263
|
|
|
super().__init__(num_osc, type_conn, type_conn_represent);
|
|
264
|
|
|
|
|
265
|
|
|
# initial states
|
|
266
|
|
|
self._excitatory = [ random.random() for _ in range(self._num_osc) ];
|
|
267
|
|
|
self._inhibitory = [0.0] * self._num_osc;
|
|
268
|
|
|
self._potential = [0.0] * self._num_osc;
|
|
269
|
|
|
|
|
270
|
|
|
self._coupling_term = None; # coupling term of each oscillator
|
|
271
|
|
|
self._global_inhibitor = 0; # value of global inhibitory
|
|
272
|
|
|
self._stimulus = None; # stimulus of each oscillator
|
|
273
|
|
|
|
|
274
|
|
|
self._dynamic_coupling = None; # dynamic connection between oscillators
|
|
275
|
|
|
self._coupling_term = [0.0] * self._num_osc;
|
|
276
|
|
|
self._buffer_coupling_term = [0.0] * self._num_osc;
|
|
277
|
|
|
|
|
278
|
|
|
# generate first noises
|
|
279
|
|
|
self._noise = [random.random() * self._params.ro for i in range(self._num_osc)];
|
|
280
|
|
|
|
|
281
|
|
|
|
|
282
|
|
|
def __del__(self):
|
|
283
|
|
|
"""!
|
|
284
|
|
|
@brief Default destructor of LEGION.
|
|
285
|
|
|
|
|
286
|
|
|
"""
|
|
287
|
|
|
if (self.__ccore_legion_pointer is not None):
|
|
288
|
|
|
wrapper.legion_destroy(self.__ccore_legion_pointer);
|
|
289
|
|
|
self.__ccore_legion_pointer = None;
|
|
290
|
|
|
|
|
291
|
|
|
|
|
292
|
|
|
def __len__(self):
|
|
293
|
|
|
"""!
|
|
294
|
|
|
@brief (uint) Returns size of LEGION.
|
|
295
|
|
|
|
|
296
|
|
|
"""
|
|
297
|
|
|
|
|
298
|
|
|
if (self.__ccore_legion_pointer is not None):
|
|
299
|
|
|
return wrapper.legion_get_size(self.__ccore_legion_pointer);
|
|
300
|
|
|
|
|
301
|
|
|
return self._num_osc;
|
|
302
|
|
|
|
|
303
|
|
|
|
|
304
|
|
|
def __create_stimulus(self, stimulus):
|
|
305
|
|
|
"""!
|
|
306
|
|
|
@brief Create stimulus for oscillators in line with stimulus map and parameters.
|
|
307
|
|
|
|
|
308
|
|
|
@param[in] stimulus (list): Stimulus for oscillators that is represented by list, number of stimulus should be equal number of oscillators.
|
|
309
|
|
|
|
|
310
|
|
|
"""
|
|
311
|
|
|
|
|
312
|
|
|
if (len(stimulus) != self._num_osc):
|
|
313
|
|
|
raise NameError("Number of stimulus should be equal number of oscillators in the network.");
|
|
314
|
|
|
else:
|
|
315
|
|
|
self._stimulus = [];
|
|
316
|
|
|
|
|
317
|
|
|
for val in stimulus:
|
|
318
|
|
|
if (val > 0): self._stimulus.append(self._params.I);
|
|
319
|
|
|
else: self._stimulus.append(0);
|
|
320
|
|
|
|
|
321
|
|
|
|
|
322
|
|
|
def __create_dynamic_connections(self):
|
|
323
|
|
|
"""!
|
|
324
|
|
|
@brief Create dynamic connection in line with input stimulus.
|
|
325
|
|
|
|
|
326
|
|
|
"""
|
|
327
|
|
|
|
|
328
|
|
|
if (self._stimulus is None):
|
|
329
|
|
|
raise NameError("Stimulus should initialed before creation of the dynamic connections in the network.");
|
|
330
|
|
|
|
|
331
|
|
|
self._dynamic_coupling = [ [0] * self._num_osc for i in range(self._num_osc)];
|
|
332
|
|
|
|
|
333
|
|
|
for i in range(self._num_osc):
|
|
334
|
|
|
neighbors = self.get_neighbors(i);
|
|
335
|
|
|
|
|
336
|
|
|
if ( (len(neighbors) > 0) and (self._stimulus[i] > 0) ):
|
|
337
|
|
|
number_stimulated_neighbors = 0.0;
|
|
338
|
|
|
for j in neighbors:
|
|
339
|
|
|
if (self._stimulus[j] > 0):
|
|
340
|
|
|
number_stimulated_neighbors += 1.0;
|
|
341
|
|
|
|
|
342
|
|
|
if (number_stimulated_neighbors > 0):
|
|
343
|
|
|
dynamic_weight = self._params.Wt / number_stimulated_neighbors;
|
|
344
|
|
|
|
|
345
|
|
|
for j in neighbors:
|
|
346
|
|
|
self._dynamic_coupling[i][j] = dynamic_weight;
|
|
347
|
|
|
|
|
348
|
|
|
|
|
349
|
|
|
def simulate(self, steps, time, stimulus, solution = solve_type.RK4, collect_dynamic = True):
|
|
350
|
|
|
"""!
|
|
351
|
|
|
@brief Performs static simulation of LEGION oscillatory network.
|
|
352
|
|
|
|
|
353
|
|
|
@param[in] steps (uint): Number steps of simulations during simulation.
|
|
354
|
|
|
@param[in] time (double): Time of simulation.
|
|
355
|
|
|
@param[in] stimulus (list): Stimulus for oscillators, number of stimulus should be equal to number of oscillators,
|
|
356
|
|
|
example of stimulus for 5 oscillators [0, 0, 1, 1, 0], value of stimulus is defined by parameter 'I'.
|
|
357
|
|
|
@param[in] solution (solve_type): Method that is used for differential equation.
|
|
358
|
|
|
@param[in] collect_dynamic (bool): If True - returns whole dynamic of oscillatory network, otherwise returns only last values of dynamics.
|
|
359
|
|
|
|
|
360
|
|
|
@return (list) Dynamic of oscillatory network. If argument 'collect_dynamic' = True, than return dynamic for the whole simulation time,
|
|
361
|
|
|
otherwise returns only last values (last step of simulation) of dynamic.
|
|
362
|
|
|
|
|
363
|
|
|
"""
|
|
364
|
|
|
|
|
365
|
|
|
if (self.__ccore_legion_pointer is not None):
|
|
366
|
|
|
pointer_dynamic = wrapper.legion_simulate(self.__ccore_legion_pointer, steps, time, solution, collect_dynamic, stimulus);
|
|
367
|
|
|
return legion_dynamic(None, None, None, pointer_dynamic);
|
|
368
|
|
|
|
|
369
|
|
|
# Check solver before simulation
|
|
370
|
|
|
if (solution == solve_type.FAST):
|
|
371
|
|
|
raise NameError("Solver FAST is not support due to low accuracy that leads to huge error.");
|
|
372
|
|
|
|
|
373
|
|
|
elif (solution == solve_type.RKF45):
|
|
374
|
|
|
raise NameError("Solver RKF45 is not support in python version. RKF45 is supported in CCORE implementation.");
|
|
375
|
|
|
|
|
376
|
|
|
# set stimulus
|
|
377
|
|
|
self.__create_stimulus(stimulus);
|
|
378
|
|
|
|
|
379
|
|
|
# calculate dynamic weights
|
|
380
|
|
|
self.__create_dynamic_connections();
|
|
381
|
|
|
|
|
382
|
|
|
dyn_exc = None;
|
|
383
|
|
|
dyn_time = None;
|
|
384
|
|
|
dyn_ginh = None;
|
|
385
|
|
|
|
|
386
|
|
|
# Store only excitatory of the oscillator
|
|
387
|
|
|
if (collect_dynamic == True):
|
|
388
|
|
|
dyn_exc = [];
|
|
389
|
|
|
dyn_time = [];
|
|
390
|
|
|
dyn_ginh = [];
|
|
391
|
|
|
|
|
392
|
|
|
step = time / steps;
|
|
393
|
|
|
int_step = step / 10.0;
|
|
394
|
|
|
|
|
395
|
|
|
for t in numpy.arange(step, time + step, step):
|
|
396
|
|
|
# update states of oscillators
|
|
397
|
|
|
self._calculate_states(solution, t, step, int_step);
|
|
398
|
|
|
|
|
399
|
|
|
# update states of oscillators
|
|
400
|
|
|
if (collect_dynamic == True):
|
|
401
|
|
|
dyn_exc.append(self._excitatory);
|
|
402
|
|
|
dyn_time.append(t);
|
|
403
|
|
|
dyn_ginh.append(self._global_inhibitor);
|
|
404
|
|
|
else:
|
|
405
|
|
|
dyn_exc = self._excitatory;
|
|
406
|
|
|
dyn_time = t;
|
|
407
|
|
|
dyn_ginh = self._global_inhibitor;
|
|
408
|
|
|
|
|
409
|
|
|
return legion_dynamic(dyn_exc, dyn_ginh, dyn_time);
|
|
410
|
|
|
|
|
411
|
|
|
|
|
412
|
|
|
def _calculate_states(self, solution, t, step, int_step):
|
|
|
|
|
|
|
413
|
|
|
"""!
|
|
414
|
|
|
@brief Calculates new state of each oscillator in the network.
|
|
415
|
|
|
|
|
416
|
|
|
@param[in] solution (solve_type): Type solver of the differential equation.
|
|
417
|
|
|
@param[in] t (double): Current time of simulation.
|
|
418
|
|
|
@param[in] step (double): Step of solution at the end of which states of oscillators should be calculated.
|
|
419
|
|
|
@param[in] int_step (double): Step differentiation that is used for solving differential equation.
|
|
420
|
|
|
|
|
421
|
|
|
"""
|
|
422
|
|
|
|
|
423
|
|
|
next_excitatory = [0.0] * self._num_osc;
|
|
424
|
|
|
next_inhibitory = [0.0] * self._num_osc;
|
|
425
|
|
|
|
|
426
|
|
|
next_potential = [];
|
|
427
|
|
|
if (self._params.ENABLE_POTENTIONAL is True):
|
|
428
|
|
|
next_potential = [0.0] * self._num_osc;
|
|
429
|
|
|
|
|
430
|
|
|
# Update states of oscillators
|
|
431
|
|
|
for index in range (0, self._num_osc, 1):
|
|
432
|
|
|
if (self._params.ENABLE_POTENTIONAL is True):
|
|
433
|
|
|
result = odeint(self._legion_state, [self._excitatory[index], self._inhibitory[index], self._potential[index]], numpy.arange(t - step, t, int_step), (index , ));
|
|
434
|
|
|
[ next_excitatory[index], next_inhibitory[index], next_potential[index] ] = result[len(result) - 1][0:3];
|
|
435
|
|
|
|
|
436
|
|
|
else:
|
|
437
|
|
|
result = odeint(self._legion_state_simplify, [self._excitatory[index], self._inhibitory[index] ], numpy.arange(t - step, t, int_step), (index , ));
|
|
438
|
|
|
[ next_excitatory[index], next_inhibitory[index] ] = result[len(result) - 1][0:2];
|
|
439
|
|
|
|
|
440
|
|
|
# Update coupling term
|
|
441
|
|
|
neighbors = self.get_neighbors(index);
|
|
442
|
|
|
|
|
443
|
|
|
coupling = 0
|
|
444
|
|
|
for index_neighbor in neighbors:
|
|
445
|
|
|
coupling += self._dynamic_coupling[index][index_neighbor] * heaviside(self._excitatory[index_neighbor] - self._params.teta_x);
|
|
446
|
|
|
|
|
447
|
|
|
self._buffer_coupling_term[index] = coupling - self._params.Wz * heaviside(self._global_inhibitor - self._params.teta_xz);
|
|
448
|
|
|
|
|
449
|
|
|
# Update state of global inhibitory
|
|
450
|
|
|
result = odeint(self._global_inhibitor_state, self._global_inhibitor, numpy.arange(t - step, t, int_step), (None, ));
|
|
451
|
|
|
self._global_inhibitor = result[len(result) - 1][0];
|
|
452
|
|
|
|
|
453
|
|
|
self._noise = [random.random() * self._params.ro for i in range(self._num_osc)];
|
|
454
|
|
|
self._coupling_term = self._buffer_coupling_term[:];
|
|
455
|
|
|
self._inhibitory = next_inhibitory[:];
|
|
456
|
|
|
self._excitatory = next_excitatory[:];
|
|
457
|
|
|
|
|
458
|
|
|
if (self._params.ENABLE_POTENTIONAL is True):
|
|
459
|
|
|
self._potential = next_potential[:];
|
|
460
|
|
|
|
|
461
|
|
|
|
|
462
|
|
|
|
|
463
|
|
|
def _global_inhibitor_state(self, z, t, argv):
|
|
|
|
|
|
|
464
|
|
|
"""!
|
|
465
|
|
|
@brief Returns new value of global inhibitory
|
|
466
|
|
|
|
|
467
|
|
|
@param[in] z (dobule): Current value of inhibitory.
|
|
468
|
|
|
@param[in] t (double): Current time of simulation.
|
|
469
|
|
|
@param[in] argv (tuple): It's not used, can be ignored.
|
|
470
|
|
|
|
|
471
|
|
|
@return (double) New value if global inhibitory (not assign).
|
|
472
|
|
|
|
|
473
|
|
|
"""
|
|
474
|
|
|
|
|
475
|
|
|
sigma = 0.0;
|
|
476
|
|
|
|
|
477
|
|
|
for x in self._excitatory:
|
|
478
|
|
|
if (x > self._params.teta_zx):
|
|
479
|
|
|
sigma = 1.0;
|
|
480
|
|
|
break;
|
|
481
|
|
|
|
|
482
|
|
|
return self._params.fi * (sigma - z);
|
|
483
|
|
|
|
|
484
|
|
|
|
|
485
|
|
|
def _legion_state_simplify(self, inputs, t, argv):
|
|
|
|
|
|
|
486
|
|
|
"""!
|
|
487
|
|
|
@brief Returns new values of excitatory and inhibitory parts of oscillator of oscillator.
|
|
488
|
|
|
@details Simplify model doesn't consider oscillator potential.
|
|
489
|
|
|
|
|
490
|
|
|
@param[in] inputs (list): Initial values (current) of oscillator [excitatory, inhibitory].
|
|
491
|
|
|
@param[in] t (double): Current time of simulation.
|
|
492
|
|
|
@param[in] argv (uint): Extra arguments that are not used for integration - index of oscillator.
|
|
493
|
|
|
|
|
494
|
|
|
@return (list) New values of excitatoty and inhibitory part of oscillator (not assign).
|
|
495
|
|
|
|
|
496
|
|
|
"""
|
|
497
|
|
|
|
|
498
|
|
|
index = argv;
|
|
499
|
|
|
|
|
500
|
|
|
x = inputs[0]; # excitatory
|
|
501
|
|
|
y = inputs[1]; # inhibitory
|
|
502
|
|
|
|
|
503
|
|
|
dx = 3.0 * x - x ** 3.0 + 2.0 - y + self._stimulus[index] + self._coupling_term[index] + self._noise[index];
|
|
504
|
|
|
dy = self._params.eps * (self._params.gamma * (1.0 + math.tanh(x / self._params.betta)) - y);
|
|
505
|
|
|
|
|
506
|
|
|
neighbors = self.get_neighbors(index);
|
|
507
|
|
|
potential = 0.0;
|
|
|
|
|
|
|
508
|
|
|
|
|
509
|
|
|
for index_neighbor in neighbors:
|
|
510
|
|
|
potential += self._params.T * heaviside(self._excitatory[index_neighbor] - self._params.teta_x);
|
|
511
|
|
|
|
|
512
|
|
|
return [dx, dy];
|
|
513
|
|
|
|
|
514
|
|
|
|
|
515
|
|
|
def _legion_state(self, inputs, t, argv):
|
|
516
|
|
|
"""!
|
|
517
|
|
|
@brief Returns new values of excitatory and inhibitory parts of oscillator and potential of oscillator.
|
|
518
|
|
|
|
|
519
|
|
|
@param[in] inputs (list): Initial values (current) of oscillator [excitatory, inhibitory, potential].
|
|
520
|
|
|
@param[in] t (double): Current time of simulation.
|
|
521
|
|
|
@param[in] argv (uint): Extra arguments that are not used for integration - index of oscillator.
|
|
522
|
|
|
|
|
523
|
|
|
@return (list) New values of excitatoty and inhibitory part of oscillator and new value of potential (not assign).
|
|
524
|
|
|
|
|
525
|
|
|
"""
|
|
526
|
|
|
|
|
527
|
|
|
index = argv;
|
|
528
|
|
|
|
|
529
|
|
|
x = inputs[0]; # excitatory
|
|
530
|
|
|
y = inputs[1]; # inhibitory
|
|
531
|
|
|
p = inputs[2]; # potential
|
|
532
|
|
|
|
|
533
|
|
|
potential_influence = heaviside(p + math.exp(-self._params.alpha * t) - self._params.teta);
|
|
534
|
|
|
|
|
535
|
|
|
dx = 3.0 * x - x ** 3.0 + 2.0 - y + self._stimulus[index] * potential_influence + self._coupling_term[index] + self._noise[index];
|
|
536
|
|
|
dy = self._params.eps * (self._params.gamma * (1.0 + math.tanh(x / self._params.betta)) - y);
|
|
537
|
|
|
|
|
538
|
|
|
neighbors = self.get_neighbors(index);
|
|
539
|
|
|
potential = 0.0;
|
|
540
|
|
|
|
|
541
|
|
|
for index_neighbor in neighbors:
|
|
542
|
|
|
potential += self._params.T * heaviside(self._excitatory[index_neighbor] - self._params.teta_x);
|
|
543
|
|
|
|
|
544
|
|
|
dp = self._params.lamda * (1.0 - p) * heaviside(potential - self._params.teta_p) - self._params.mu * p;
|
|
545
|
|
|
|
|
546
|
|
|
return [dx, dy, dp]; |
annoviko /
pyclustering
Loading history...
This can be caused by one of the following:
1. Missing Dependencies
This error could indicate a configuration issue of Pylint. Make sure that your libraries are available by adding the necessary commands.
2. Missing __init__.py files
This error could also result from missing
__init__.pyfiles in your module folders. Make sure that you place one file in each sub-folder.