def visited(self):
"""
Return a list of visited nodes.
"""
# assert nodes have been visited
assert self.r.sum() != 0
x = np.where(self.r > 0)[1]
visited = remove_array_duplicates(x)
return visited
def visited(self):
"""
Return a list of visited nodes.
"""
# assert nodes have been visited
assert self.r.sum() != 0
x = np.where(self.r > 0)[1]
visited = remove_array_duplicates(x)
return visited
# wrong response so simulation runs a bit longer
# which leaves extra space after the winner
# has been picked
problem = ['river', 'note', 'account']
target = 'house'
param = {'max_visited': 5, 'stim_len': 50}
# run the simulation
network = Network(**param)
network.setup_problem(problem, target)
network.run()
# get indices of winning units
winners = np.where(network.w > network.theta_w)[1]
unique_win = remove_array_duplicates(winners)
for i in unique_win:
print network.ids[i], network.a[network.t_max, i]
# get activated units which are not winners
background = np.where(network.a > 0.1)[1]
background = set(background)
ids = list(background)
idswc = list(set(ids) - set(network.cue_ids))
p, n = len(unique_win), network.t_max
X = np.linspace(0, 2, n)
Ya = network.a[:n, ids].T
# which leaves extra space after the winner
# has been picked
problem = ['river', 'note', 'account']
target = 'house'
param = {'max_visited': 5,
'stim_len': 50}
# run the simulation
network = Network(**param)
network.setup_problem(problem, target)
network.run()
# get indices of winning units
winners = np.where(network.w > network.theta_w)[1]
unique_win = remove_array_duplicates(winners)
for i in unique_win:
print network.ids[i], network.a[network.t_max, i]
# get activated units which are not winners
background = np.where(network.a > 0.1)[1]
background = set(background)
ids = list(background)
idswc = list(set(ids) - set(network.cue_ids))
p, n = len(unique_win), network.t_max
X = np.linspace(0, 2, n)
Ya = network.a[:n, ids].T
