-
Notifications
You must be signed in to change notification settings - Fork 1
/
SEM_ring_wta_current_input.py
461 lines (395 loc) · 16.6 KB
/
SEM_ring_wta_current_input.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
import sys, copy
from nest import *
import nest.topology as tp
import numpy as np
from pylab import cm
import matplotlib.pyplot as plt
import pickle
# This package generates 28 by 28 pixel images
# It can also generate the input poisson spikes directly
# Has to be tested whether this is faster than the Pynn Poisson sources
import SEM_input
last_events = [0,0,0,0]
results_dir = './results'
with_plot_weights = False
with_homeostatis = False
seed = 764756387
tshow = 40.0 # ms
tpaus = 10.0
input_rate = 40.0 # Hz
cell_params_lif_cond = {
'V_th' :-55.0, # mV
'g_L' : 10.0, # ms
'tau_syn_ex' : 2.0, # ms
'tau_syn_in' : 3.0, # ms
'tau_minus' : 20.0,
't_ref':5.0,
}
cell_params_lif = {
'tau_m' : 5.0, # ms
}
# values taken from Naud et al. 2008
cell_params_adex = {
'C_m' : 200.0, # pF
'g_L' : 10.0, # ms
#'tau_refrac': 2.0, # ms
'V_th' :-50.0, # mV
#'tau_syn_E' : 2.5, # ms
#'tau_syn_I' : 5.0, # ms
'E_L' :-58.0, # mV
'a' : 2.0, # nS
'b' :100.0, # pA
# V_T and delta_t default OK
'tau_w' :120.0, # ms
'V_reset' :-46.0, # mV
'tau_syn_ex': 1.0,
'tau_syn_in': 1.5,
'gsl_error_tol': 1e-8,
}
#SetDefaults('iaf_cond_exp',cell_params_lif_cond)
SetDefaults('iaf_neuron',cell_params_lif)
#SetDefaults('aeif_cond_exp',cell_params_adex)
import pprint
pp = pprint.PrettyPrinter()
pp.pprint(GetDefaults('iaf_neuron'))
pp.pprint(GetDefaults('aeif_cond_exp'))
#pp.pprint(GetDefaults('aeif_cond_exp'))
#print GetStatus([0])
SetStatus([0],{'overwrite_files':True, 'data_path':results_dir})
#sys.exit(0)
num = {}
num['l0_exc_neurons'] = 4
num['l0_exc_maxneighbors'] = 4
num['l1_exc_neurons'] = 4
num['l1_exc_maxneighbors'] = 4
num['l0_l1_maxneighbors'] = 4
num['l0_inh_neurons'] = 4
num['l1_inh_neurons'] = 5
num['inputs'] = SEM_input.SEM_input_config['num_inputs']
num['inputs_inh_sources'] = 4
num['steps'] = 5000
num['steps_rate_average'] = 20
# input -> exc0
w_inp_exc0_peak = 180.0
sigma_inp_exc0 = num['inputs']/3.0
w_inp_exc0_max = 250.0
alpham_inp_exc0 = 0.01/200
alphap_inp_exc0 = alpham_inp_exc0*2
w_inp_exc0_min = 1.0
# input -> inh0
w_inp_inh0 = 150.0
sigma_inp_inh0 = num['inputs']/2.0
# exc0 -> exc1
sigma_exc0_exc1 = 4.
w_exc0_exc1_peak = 0.005
w_exc0_exc1_max = 0.005
# exc0 -> inh0
p_exc0_inh0 = 10.0
w_exc0_inh0 = 1000.0
# exc0 -> exc0
sigma_exc0_exc0 = 0.5
w_exc0_exc0_max = 0.5
# exc1 -> exc1
sigma_exc1_exc1 = 7.
w_exc1_exc1_max = 0.001
# exc1 -> inh1
p_exc1_inh1 = 1.0
w_exc1_inh1 = 0.02
# inh0 -> exc0
p_inh0_exc0 = 1.0
w_inh0_exc0 = -1000.0
# inh1 -> exc1
p_inh1_exc1 = 1.0
w_inh1_exc1 = 0.1
def plot_neurons(meters,k,m,variables,title):
plt.figure()
plt.suptitle(title)
start = len(variables)*100+11
for var in variables:
plt.subplot(start)
plt.grid()
for i in range(k,m):
#meter = tp.GetElement(meters,[i,0])
events = GetStatus([meters[i]])[0]['events']
t = events['times']
plt.plot(t, events[var])
plt.ylabel(var)
start = start + 1
def plot_weights(layer,target):
image_width = SEM_input.SEM_input_config['image_width']
weights = np.zeros((image_width,image_width))
for i in range(0,image_width):
for j in range(0,image_width):
pixel = (i*image_width+j)*2
status = GetStatus(FindConnections(tp.GetElement(layer,[pixel,0])))
for s in status:
if s['target'] == target:
weights[i][j] = s['weight']
plt.figure()
#plt.hist(weights,bins=100)
#cmap = plt.get_cmap('grey')
plt.title(str(target))
plt.imshow(weights)
plt.colorbar()
def dump_weights(layer,tag=''):
layer_status = GetStatus(layer)[0]
image_width = SEM_input.SEM_input_config['image_width']
weights = []
for i in range(0,image_width):
for j in range(0,image_width):
pixel = i*image_width+j
status = GetStatus(FindConnections(tp.GetElement(layer,[pixel,0])))
for s in status:
weights.append((i,j,s['target'],s['weight']))
pickle.dump(weights,open(results_dir+'/weights_'+str(layer_status['global_id'])+'_'+tag+'.dat','w'))
def plot_layer(layer):
tp.PlotLayer(layer, nodesize=50)
status = GetStatus(layer)[0]
extent = status['topology']['extent']
# beautify
plt.axis([extent[0]/(-2.0)-0.25, extent[0]/2.0+0.25, extent[1]/(-2.0)-0.25, extent[1]/2.0+0.25])
plt.axes().set_aspect('equal', 'box')
plt.axes().set_xticks([w-extent[0]/2.0 for w in range(int(extent[0])+1)])
plt.axes().set_yticks([w-extent[1]/2.0 for w in range(int(extent[1])+1)])
plt.grid(True)
plt.xlabel('%d Columns, Extent: %f' % (status['topology']['columns'],extent[0]))
plt.ylabel('%d Rows, Extent: %f' % (status['topology']['rows'],extent[1]))
def setup_network():
print "[ Creating excitatory population ]"
# 1) create excitatory populations
l0_exc_population = tp.CreateLayer ({
'rows' : 1 ,
'columns' : num['l0_exc_neurons'],
'extent' : [ float(num['l0_exc_neurons']), 1. ],
'elements' : 'iaf_neuron',
'edge_wrap' : True } )
print "[ Creating inhibitory population ]"
# 2) create inhibitory population
l0_inh_population = tp.CreateLayer ({
'rows' : 1,
'columns' : num['l0_inh_neurons'],
'extent' : [ float(num['l0_inh_neurons']), 1. ] ,
'elements' : 'iaf_neuron',
'edge_wrap' : True } )
print "[ Projecting excitatory -> inhibitory population ]"
# 3) exc -> inh
CopyModel('static_synapse','exc_inh', {'weight':w_exc0_inh0})
l0_exc_inh_dict = {
'connection_type' : 'convergent',
'mask': {
'rectangular': {
'lower_left': [-0.5,-0.5],
'upper_right':[0.5,0.5]
}
},
#'kernel':p_exc0_inh0,
'synapse_model': 'exc_inh'
}
tp.ConnectLayers(l0_exc_population,l0_inh_population,l0_exc_inh_dict)
print "[ Projecting excitatory -> excitatory population ]"
# 4) exc -> exc
# Since these connections are distance-dependent, we can use topological connections here
CopyModel('static_synapse','exc_exc',{'weight':w_exc0_exc0_max})
l0_exc_exc_dict = {
'connection_type': 'convergent',
'mask': {
'rectangular': {
'lower_left': [float(num['l0_exc_neurons']/(-2.0)),-0.5],
'upper_right':[float(num['l0_exc_neurons']/(2.0)),0.5]
}
},
'synapse_model': 'exc_exc',
'weights': {
'gaussian': {'p_center': 1., 'sigma': sigma_exc0_exc0}
}
}
#tp.ConnectLayers(l0_exc_population,l0_exc_population,l0_exc_exc_dict)
print "[ Projecting inhibitory -> excitatory population ]"
# 5) inh -> exc
CopyModel('static_synapse','inh_exc', {'weight':w_inh0_exc0})
l0_inh_exc_dict = {
'connection_type' : 'convergent',
'mask': {
'grid': {
'rows': 1,
'columns': num['l0_inh_neurons']-1,
},
'anchor': {
'row': 0,
'column': -1,
}
},
#'kernel':p_inh0_exc0,
'synapse_model': 'inh_exc'
}
tp.ConnectLayers(l0_inh_population,l0_exc_population,l0_inh_exc_dict)
print "[ Creating input population ]"
# 6) create input population
CopyModel('poisson_generator','input_model',{'rate':0.0})
input_population = tp.CreateLayer ({
'rows' : 1 ,
'columns' : num['inputs'],
'extent' : [ float(num['inputs']), 1. ],
'elements' : 'input_model',
'edge_wrap' : True } )
parrot_population = tp.CreateLayer ({
'rows' : 1 ,
'columns' : num['inputs'],
'extent' : [ float(num['inputs']), 1. ],
'elements' : 'parrot_neuron',
'edge_wrap' : True } )
print "[ Projecting input -> excitatory population ]"
# 7) input -> exc
CopyModel('static_synapse','input_parrot', {'weight': 1.0})
CopyModel('rect_stdp_synapse','parrot_exc', {'tau_plus': 10.0,'alpha_minus':alpham_inp_exc0,'alpha_plus':alphap_inp_exc0,'Wmax':w_inp_exc0_max})
#CopyModel('static_synapse','parrot_exc')
CopyModel('static_synapse','parrot_inh', {'weight': w_inp_inh0})
input_parrot_dict = {
'connection_type' : 'convergent',
'mask': {
'rectangular': {
# this is intended: only connect to one neuron!
'lower_left': [-0.5,-0.5],
'upper_right':[0.5,0.5]
}
},
'synapse_model': 'input_parrot'
}
parrot_exc_dict = {
'connection_type' : 'convergent',
'mask': {
'grid': {
'rows': 1,
'columns': num['inputs'],
},
'anchor': {
'row': 0,
'column': 0,
}
},
#'kernel': {'gaussian': {'p_center':1.0,'sigma':sigma_inp_exc0}},
'weights': {'uniform':{'min':w_inp_exc0_min,'max':w_inp_exc0_peak}},
'synapse_model': 'parrot_exc'
}
parrot_inh_dict = copy.copy(parrot_exc_dict)
#parrot_inh_dict['synapse_model'] = 'parrot_inh'
#parrot_inh_dict['kernel'] = {'gaussian': {'p_center':1.0,'sigma':sigma_inp_exc0}}
tp.ConnectLayers(input_population,parrot_population,input_parrot_dict)
tp.ConnectLayers(parrot_population,l0_exc_population,parrot_exc_dict)
#tp.ConnectLayers(parrot_population,l0_inh_population,parrot_inh_dict)
# 7) input -> inh
CopyModel('poisson_generator','inh_bias_model',{'rate':500.0})
inh_bias_population = tp.CreateLayer ({
'rows' : 1 ,
'columns' : num['inputs_inh_sources'],
'extent' : [ float(num['inputs_inh_sources']), 1. ],
'elements' : 'inh_bias_model',
'edge_wrap' : True } )
CopyModel('static_synapse','bias_inh', {'weight': w_inp_inh0})
inh_bias_dict = {
'connection_type' : 'convergent',
'mask': {
'rectangular': {
# this is intended: only connect to one neuron!
'lower_left': [-0.5,-0.5],
'upper_right':[0.5,0.5]
}
},
'synapse_model': 'bias_inh',
'weights': w_inp_inh0,
}
tp.ConnectLayers(inh_bias_population,l0_inh_population,inh_bias_dict)
return l0_exc_population,l0_inh_population,input_population,parrot_population
# CREATE
populations = setup_network()
#plot_layer(populations[2])
#plot_layer(populations[3])
#plt.figure() #ugly stack cleaning
#print GetStatus(populations[0])
#print GetStatus(populations[2])
#for i in range(num['l0_exc_neurons']):
# print GetStatus(tp.GetElement(populations[0],[i,0]))
#for i in range(num['l0_inh_neurons']):
# print GetStatus(tp.GetElement(populations[1],[i,0]))
# tgts = GetStatus(FindConnections(tp.GetElement(populations[1],[i,0])))
# print tgts
#connPlot(populations[2],'input_model','aeif_cond_exp','input_exc','title')
if with_plot_weights:
plot_weights(populations[3],2)
plot_weights(populations[3],3)
plot_weights(populations[3],4)
plot_weights(populations[3],5)
# CONNECT READOUTS
# voltmeters
exc_voltmeters = Create('multimeter',num['l0_exc_neurons'],{'to_file':True,'record_from':['V_m'],'to_memory':False})
inh_voltmeters = Create('multimeter',num['l0_inh_neurons'],{'to_file':True,'record_from':['V_m'],'to_memory':False})
tgts = [nd for nd in GetLeaves(populations[0])[0]]
for i in range(num['l0_exc_neurons']):
SetStatus([exc_voltmeters[i]],{'label':'mm_exc_'+str(i)})
DivergentConnect([exc_voltmeters[i]],[tgts[i]])
tgts = [nd for nd in GetLeaves(populations[1])[0]]
for i in range(num['l0_inh_neurons']):
SetStatus([inh_voltmeters[i]],{'label':'mm_inh_'+str(i)})
DivergentConnect([inh_voltmeters[i]],[tgts[i]])
# spike detectors
exc_spikedetectors = Create('spike_detector',num['l0_exc_neurons'],{'precise_times':True,'to_file':True,'to_memory':True})
exc_tgts = [nd for nd in GetLeaves(populations[0])[0]]
#for t in exc_tgts:
# print GetStatus([t])
for n in range(num['l0_exc_neurons']):
SetStatus([exc_spikedetectors[n]],{'label':"sd_exc_"+str(n)})
ConvergentConnect([exc_tgts[n]],[exc_spikedetectors[n]])
inh_spikedetectors = Create('spike_detector',num['l0_inh_neurons'],{'precise_times':True,'to_file':True,'to_memory':False})
inh_tgts = [nd for nd in GetLeaves(populations[1])[0]]
for n in range(num['l0_inh_neurons']):
SetStatus([inh_spikedetectors[n]],{'label':"sd_inh_"+str(n)})
ConvergentConnect([inh_tgts[n]],[inh_spikedetectors[n]])
inp_spikedetectors = Create('spike_detector',1,{'precise_times':True,'to_file':True})
inp_tgts = [nd for nd in GetLeaves(populations[3])[0]]
SetStatus([inp_spikedetectors[0]],{'label':"sd_inp_"+str(0)})
for n in range(num['inputs']):
ConvergentConnect([inp_tgts[n]],[inp_spikedetectors[0]])
# DUMP INITIAL WEIGHTS
dump_weights(populations[3],'pre')
# GENERATE INPUT AND SIMULATE
inputs = [nd for nd in GetLeaves(populations[2])[0]]
for i in range(num['steps']):
image = SEM_input.draw_image(SEM_input.SEM_input_config['centers'][i%4],SEM_input.SEM_input_config)
# prepare input population firing rates
for j in range(SEM_input.SEM_input_config['image_width']):
for k in range(SEM_input.SEM_input_config['image_width']):
this_rate = SEM_input.SEM_input_config['input_on_rate'] if (image[j][k] > 0.0) else SEM_input.SEM_input_config['input_off_rate']
SetStatus([inputs[j*SEM_input.SEM_input_config['image_width']+k]],{'rate': this_rate})
#SetStatus([inputs[2*j*SEM_input.SEM_input_config['image_width']+k*2+1]],{'rate': SEM_input.SEM_input_config['input_on_rate'] if (image[j][k]<1.0) else SEM_input.SEM_input_config['input_off_rate']})
print "[Running simulation step %d]" % (i)
Simulate(tshow)
for j in range(SEM_input.SEM_input_config['image_width']):
for k in range(SEM_input.SEM_input_config['image_width']):
SetStatus([inputs[j*SEM_input.SEM_input_config['image_width']+k]],{'rate': 0.0})
#SetStatus([inputs[2*j*SEM_input.SEM_input_config['image_width']+k*2+1]],{'rate': 0.0})
Simulate(tpaus)
if i%num['steps_rate_average']==0 and i>0 and with_homeostatis:
for neuron in range(num['l0_exc_neurons']):
events_ex = GetStatus([exc_spikedetectors[neuron]],'n_events')[0]
rate = (events_ex-last_events[neuron])/((tshow+tpaus)*num['steps_rate_average']*1e-3)
last_events[neuron] = events_ex
g_L = 5.0/(1+(rate-20.0)/100.0)
print rate,g_L
SetStatus([GetLeaves(populations[0])[0][neuron]],{'tau_m':g_L})
# #if rate < 15.0:
# # SetStatus([GetLeaves(populations[0])[0][neuron]],{'g_L':5.0})
# #elif rate > 35.0:
# # SetStatus([GetLeaves(populations[0])[0][neuron]],{'g_L':15.0})
# #else:
# # SetStatus([GetLeaves(populations[0])[0][neuron]],{'g_L':10.0})
if i%100==0 and i>0:
dump_weights(populations[3],'post')
if with_plot_weights:
plot_weights(populations[3],2)
plot_weights(populations[3],3)
plot_weights(populations[3],4)
plot_weights(populations[3],5)
plt.show()
dump_weights(populations[3],'post')
# PRINT
#PrintNetwork(depth=2)