def test_neurons_to_node(Simulator, nl_nodirect):
name = 'neurons_to_node'
N = 30
m = nengo.Model(name, seed=123)
a = nengo.Ensemble(nl_nodirect(N), dimensions=1)
out = nengo.Node(lambda t, x: x, size_in=N)
nengo.Connection(a.neurons, out, filter=None)
a_spikes = nengo.Probe(a, 'spikes')
out_p = nengo.Probe(out, 'output')
sim = Simulator(m)
sim.run(0.6)
t = sim.trange()
with Plotter(Simulator, nl_nodirect) as plt:
ax = plt.subplot(111)
try:
from nengo.matplotlib import rasterplot
rasterplot(t, sim.data(a_spikes), ax=ax)
rasterplot(t, sim.data(out_p), ax=ax)
except ImportError:
pass
plt.savefig('test_connection.test_' + name + '.pdf')
plt.close()
assert np.allclose(sim.data(a_spikes)[:-1], sim.data(out_p)[1:])
def test_neurons_to_node(Simulator, nl_nodirect):
name = 'neurons_to_node'
N = 30
m = nengo.Network(label=name, seed=123)
with m:
m.config[nengo.Ensemble].neuron_type = nl_nodirect()
a = nengo.Ensemble(N, dimensions=1)
out = nengo.Node(lambda t, x: x, size_in=N)
nengo.Connection(a.neurons, out, synapse=None)
a_spikes = nengo.Probe(a.neurons)
out_p = nengo.Probe(out)
sim = Simulator(m)
sim.run(0.6)
t = sim.trange()
with Plotter(Simulator, nl_nodirect) as plt:
ax = plt.subplot(111)
try:
from nengo.matplotlib import rasterplot
rasterplot(t, sim.data[a_spikes], ax=ax)
rasterplot(t, sim.data[out_p], ax=ax)
except ImportError:
pass
plt.savefig('test_connection.test_' + name + '.pdf')
plt.close()
assert np.allclose(sim.data[a_spikes], sim.data[out_p])
def test_neurons_to_node(Simulator, nl_nodirect):
name = 'neurons_to_node'
N = 30
m = nengo.Network(label=name, seed=123)
with m:
m.config[nengo.Ensemble].neuron_type = nl_nodirect()
a = nengo.Ensemble(N, dimensions=1)
out = nengo.Node(lambda t, x: x, size_in=N)
nengo.Connection(a.neurons, out, synapse=None)
a_spikes = nengo.Probe(a, 'spikes')
out_p = nengo.Probe(out, 'output')
sim = Simulator(m)
sim.run(0.6)
t = sim.trange()
with Plotter(Simulator, nl_nodirect) as plt:
ax = plt.subplot(111)
try:
from nengo.matplotlib import rasterplot
rasterplot(t, sim.data[a_spikes], ax=ax)
rasterplot(t, sim.data[out_p], ax=ax)
except ImportError:
pass
plt.savefig('test_connection.test_' + name + '.pdf')
plt.close()
assert np.allclose(sim.data[a_spikes][:-1], sim.data[out_p][1:])
def _plot(data_fname, plot_fname, params, sims, sims_time, before_spikes,
before_time, after_spikes, after_time, show=False):
print "Plotting..."
fig = plt.figure()
plt.subplots_adjust(right=0.93, top=0.98, bottom=0.11, left=0.07, hspace=0.05)
#plt.rc('text', usetex=True)
mpl.rcParams['lines.linewidth'] = 1.5
mpl.rc('font', size=14)
xticks = [2 * i * params.testing_time for i in range(4)]
time = before_time
spacing = 4
title_fontsize = 20
ax = plt.subplot(211)
ax_spike = ax.twinx()
ax.plot(sims_time, sims, color="Gray")
spikes = np.concatenate((np.zeros((time.size, spacing)), before_spikes), axis=1)
rasterplot(time, spikes, ax=ax_spike, color="Black")
ax.set_ylim((0.0, 1.05))
ax.set_ylabel(r'$Similarity$', fontsize=title_fontsize)
ax.set_yticks([0,1])
ax.set_xticks(xticks)
ax.xaxis.set_ticklabels([])
ax_spike.set_yticks([])
ax = plt.subplot(212)
ax_spike = ax.twinx()
ax.plot(sims_time, sims, color="Gray")
spikes = np.concatenate((np.zeros((time.size, spacing)), after_spikes), axis=1)
rasterplot(time, spikes, ax=ax_spike, color="Black")
ax.xaxis.set_tick_params(pad=8)
ax.set_xlabel(r'$Time\ (s)$', fontsize=title_fontsize)
ax.set_xticks(xticks)
ax.set_ylabel(r'$Similarity$', fontsize=title_fontsize)
ax.set_ylim((0.0, 1.05))
ax.set_yticks([0,1])
ax_spike.set_yticks([])
plt.savefig(plot_fname)
if show:
plt.show()
offset += 1
plt.subplot(offset)
plt.plot(t, sim.data(error_p)[-trunc:, :], label='Error')
plt.ylabel('Error: Decoded')
remove_xlabels()
offset += 1
plt.subplot(offset)
plt.plot(t, sim.data(cleanup_p)[-trunc:, :], label='Cleanup')
plt.ylabel('Cleanup: (Orig. Decoders)')
remove_xlabels()
offset += 1
plt.subplot(offset)
rasterplot(spike_time, sim.data(cleanup_s)[-trunc:, :])
plt.ylabel('Cleanup: Spikes')
offset += 1
plt.subplot(offset)
connection_weights = np.squeeze(sim.data(weights_p)[-trunc:, :, :])
maxes = np.amax(connection_weights, 1)
mins = np.amin(connection_weights, 1)
plt.plot(t, maxes, label='weights')
plt.plot(t, mins, label='weights')
plt.ylabel('Oja Weights')
offset += 1
if plot_connection_weights:
extremes_only = True
for i in range(min(1, cleanup_n)):
offset += 1
plt.subplot(offset)
plt.plot(t, sim.data(error_p)[-trunc:,:], label='Error')
plt.ylabel('Error: Decoded')
remove_xlabels()
offset += 1
plt.subplot(offset)
plt.plot(t, sim.data(cleanup_p)[-trunc:,:], label='Cleanup')
plt.ylabel('Cleanup: (Orig. Decoders)')
remove_xlabels()
offset += 1
plt.subplot(offset)
rasterplot(spike_time, sim.data(cleanup_s)[-trunc:,:])
plt.ylabel('Cleanup: Spikes')
offset += 1
plt.subplot(offset)
connection_weights = np.squeeze(sim.data(weights_p)[-trunc:,:,:])
maxes = np.amax(connection_weights, 1)
mins = np.amin(connection_weights, 1)
plt.plot(t, maxes, label='weights')
plt.plot(t, mins, label='weights')
plt.ylabel('Oja Weights')
offset += 1
if plot_connection_weights:
extremes_only = True
for i in range(min(1, cleanup_n)):
num_plots = 7
offset = num_plots * 100 + 10 + 1
# ----- Plot! -----
print "Plotting..."
#Testing Phase 1
plt.subplot(offset)
plt.plot(t1, inn1)
plt.ylabel('Input')
offset += 1
plt.title('Clean then noisy')
plt.subplot(offset)
rasterplot(t1, spikes1)
plt.ylabel('Cleanup: Spikes')
offset += 1
#Testing Phase 2
plt.subplot(offset)
plt.plot(t1, inn2[-len(t1):])
plt.ylabel('Input')
offset += 1
plt.subplot(offset)
plt.plot(t1, [np.dot(i, training_vector) for i in inn2[-len(t1):]])
plt.ylabel('Similarity to training_vector')
offset += 1
plt.subplot(offset)
num_plots = 7
offset = num_plots * 100 + 10 + 1
# ----- Plot! -----
print "Plotting..."
#Testing Phase 1
plt.subplot(offset)
plt.plot(t1, inn1)
plt.ylabel('Input')
offset += 1
plt.title('Clean then noisy')
plt.subplot(offset)
rasterplot(t1, spikes1)
plt.ylabel('Cleanup: Spikes')
offset += 1
#Testing Phase 2
plt.subplot(offset)
plt.plot(t1, inn2[-len(t1):])
plt.ylabel('Input')
offset += 1
plt.subplot(offset)
plt.plot(t1, [np.dot(i, training_vector) for i in inn2[-len(t1):]])
plt.ylabel('Similarity to training_vector')
offset += 1
plt.subplot(offset)
n_output = nengo.Probe(n, 'decoded_output', filter = 0.01) # Spikes filtered by a 10ms post-synaptic filter
sim = nengo.Simulator(model) #Create a simulator
sim.run(1) # Run it for 5 seconds
import matplotlib.pyplot as plt
# Plot the decoded output of the ensemble
t = sim.data(model.t_probe) #Get the time steps
plt.plot(t, sim.data(n_output))
plt.plot(t, sim.data(sin_p))
plt.xlim(0,1)
# Plot the spiking output of the ensemble
from nengo.matplotlib import rasterplot
plt.figure(figsize=(10,8))
plt.subplot(221)
rasterplot(t, sim.data(n_spikes), colors=[(1,0,0), (0,0,0)])
plt.xlim(0,1)
plt.yticks((1, 2), ("On neuron", "Off neuron"))
# Plot the soma voltages of the neurons
#plt.subplot(222)
#data = sim.data('Neurons.voltages')
#plt.plot(t, data[:,0]+1, 'r')
#plt.plot(t, data[:,1], 'k')
#plt.yticks(())
#plt.axis([0,1,0,2])
#plt.subplots_adjust(wspace=0.05)
plt.show()
