def draw_correlation(self, graph_key, flags=None, smooth=False,reverse=False,name=None,xlabel=None,ylabel=None,color_bar=True,color_bar_label=None,vmin=None,vmax=None):
# Gathering neurons
if flags is None:
neuron_list = xrange(len(self.neuron_list))
else:
neuron_list = self.get_neuron_id_from_flags(flags)
if reverse:
neuron_list.reverse()
# Gathering datas
data = map(lambda x : self.neuron_data[x][graph_key], neuron_list)
if smooth:
data = map(lambda x : math_tools.gaussian_kernel(x,200), data)
data = numpy.array(data)
correlation_matrix = numpy.corrcoef(data)
# Drawing the graph
fig = plt.figure()
ax = fig.add_subplot(111)
img = ax.imshow(correlation_matrix, interpolation="nearest",vmax=vmax,vmin=vmin)
if color_bar:
color_bar = plt.colorbar(img)
if color_bar_label is not None:
color_bar.set_label(color_bar_label)
if name is not None:
plt.title(name)
if xlabel is not None:
plt.xlabel(xlabel)
if ylabel is not None:
plt.ylabel(ylabel)
plt.show()
def draw_neuron_graph(self, graph_key, neuron_id = None, flags = None, r_figure=True,smooth=False):
data_list = []
names = []
if type(graph_key)==list:
figure = True
else:
figure = False
graph_key = [graph_key]
if not r_figure:
figure=False
if neuron_id is not None:
data_list = map(lambda x: self.neuron_data[neuron_id][x], graph_key)
names.extend(map(lambda x: "%s for %s" % (x, self.neuron_list[neuron_id].get_name()), graph_key))
else:
if flags is not None:
if type(flags) != list:
flags = [flags]
data = []
neurons = []
for flag in flags:
data.extend(map(lambda x: self.neuron_data[x], self.neuron_flag_dict[flag]))
neurons.extend(map(lambda x: self.neuron_list[x], self.neuron_flag_dict[flag]))
else:
data = self.neuron_data
neurons = self.neuron_list
for key in graph_key:
data_list.extend(map(lambda x: x[key], data))
names.extend(map(lambda x: "%s for %s" % (key, x.get_name()), neurons))
if smooth:
data_list = map(lambda x : math_tools.gaussian_kernel(x, 200), data_list)
smart_plot(data_list, figure=figure, names=names)
def draw_neuron_graph(self,
graph_key,
neuron_id=None,
flags=None,
r_figure=True,
smooth=False):
data_list = []
names = []
if type(graph_key) == list:
figure = True
else:
figure = False
graph_key = [graph_key]
if not r_figure:
figure = False
if neuron_id is not None:
data_list = map(lambda x: self.neuron_data[neuron_id][x],
graph_key)
names.extend(
map(
lambda x: "%s for %s" %
(x, self.neuron_list[neuron_id].get_name()), graph_key))
else:
if flags is not None:
if type(flags) != list:
flags = [flags]
data = []
neurons = []
for flag in flags:
data.extend(
map(lambda x: self.neuron_data[x],
self.neuron_flag_dict[flag]))
neurons.extend(
map(lambda x: self.neuron_list[x],
self.neuron_flag_dict[flag]))
else:
data = self.neuron_data
neurons = self.neuron_list
for key in graph_key:
data_list.extend(map(lambda x: x[key], data))
names.extend(
map(lambda x: "%s for %s" % (key, x.get_name()), neurons))
if smooth:
data_list = map(lambda x: math_tools.gaussian_kernel(x, 200),
data_list)
smart_plot(data_list, figure=figure, names=names)
def draw_correlation(self,
graph_key,
flags=None,
smooth=False,
reverse=False,
name=None,
xlabel=None,
ylabel=None,
color_bar=True,
color_bar_label=None,
vmin=None,
vmax=None):
# Gathering neurons
if flags is None:
neuron_list = xrange(len(self.neuron_list))
else:
neuron_list = self.get_neuron_id_from_flags(flags)
if reverse:
neuron_list.reverse()
# Gathering datas
data = map(lambda x: self.neuron_data[x][graph_key], neuron_list)
if smooth:
data = map(lambda x: math_tools.gaussian_kernel(x, 200), data)
data = numpy.array(data)
correlation_matrix = numpy.corrcoef(data)
# Drawing the graph
fig = plt.figure()
ax = fig.add_subplot(111)
img = ax.imshow(correlation_matrix,
interpolation="nearest",
vmax=vmax,
vmin=vmin)
if color_bar:
color_bar = plt.colorbar(img)
if color_bar_label is not None:
color_bar.set_label(color_bar_label)
if name is not None:
plt.title(name)
if xlabel is not None:
plt.xlabel(xlabel)
if ylabel is not None:
plt.ylabel(ylabel)
plt.show()
