def generator_reservoir(self):
"""
Use another function to create the reservoir and return it
"""
if self.com_size != None:
minc = self.com_size
maxc = self.com_size
else:
minc = self.minc
maxc = self.maxc
if self.homok != None:
k = self.homok
maxk = self.homok
else:
k = self.k
maxk = self.maxk
self.nx_reservoir = graphgen.uniform_weighted_directed_lfr_graph(
N=self.N,
mu=self.mu,
k=k,
maxk=maxk,
minc=minc,
maxc=maxc,
deg_exp=self.deg_exp,
temp_dir_ID=self.temp_dir_ID,
full_path=self.full_path,
weight_bounds=self.reservoir_weight_scale *
self.reservoir_weight_bounds)
return np.asarray(nx.to_numpy_matrix(self.nx_reservoir))
def generator_reservoir(self):
"""
Use another function to create the reservoir and return it
"""
self.nx_reservoir = graphgen.uniform_weighted_directed_lfr_graph(
N=self.N,
mu=self.mu,
k=self.k,
maxk=self.maxk,
minc=self.minc,
maxc=self.maxc,
deg_exp=self.deg_exp,
temp_dir_ID=self.temp_dir_ID,
full_path=self.full_path,
weight_bounds=self.reservoir_weight_scale *
np.array([self.lower_reservoir_bound, self.upper_reservoir_bound]))
reservoir = np.asarray(nx.to_numpy_matrix(self.nx_reservoir))
if self.spectral_factor != None:
reservoir = np.asmatrix(reservoir) * np.identity(self.N)
effective_spectral_radius = self.calculate_largest_eigenval(
reservoir)
print effective_spectral_radius
reservoir = reservoir / effective_spectral_radius * self.spectral_factor
return reservoir
def generate_network(N, mu, k, maxk, minc, maxc, deg_exp, weight_scale,
lower_weight_bound, upper_weight_bound, temp_dir_ID,
full_path):
return graphgen.uniform_weighted_directed_lfr_graph(
N,
mu,
k,
maxk,
minc,
maxc,
weight_bounds=weight_scale *
np.array([lower_weight_bound, upper_weight_bound]),
deg_exp=deg_exp,
temp_dir_ID=temp_dir_ID,
full_path=full_path)
def generator_reservoir(self):
"""
Use another function to create the reservoir and return it
"""
self.nx_reservoir = graphgen.uniform_weighted_directed_lfr_graph(
N=self.N,
mu=self.mu,
k=self.k,
maxk=self.maxk,
minc=self.minc,
maxc=self.maxc,
deg_exp=self.deg_exp,
temp_dir_ID=self.temp_dir_ID,
full_path=self.full_path,
weight_bounds=self.reservoir_weight_scale *
np.array([self.lower_reservoir_bound, self.upper_reservoir_bound]))
return np.asarray(nx.to_numpy_matrix(self.nx_reservoir))
import numpy as np
import networkx as nx
import graphgen
if __name__ == '__main__':
"""
"""
graph = graphgen.uniform_weighted_directed_lfr_graph(N=40,
mu=0.5,
k=5,
maxk=5,
minc=10,
maxc=10,
deg_exp=1,
temp_dir_ID=0,
full_path=None,
weight_bounds=1 *
np.array([1, 1]))
nx.write_gexf(graph, "high_mu_graph.gexf")
plt.xlabel("time")
plt.ylabel("activity")
plt.legend(loc=2)
plt.xlim(1, len(activity) + 1)
plt.tight_layout()
plt.savefig(prefix + "_com_activity_ts.png", dpi=300)
plt.clf()
plt.close()
nx_reservoir = graphgen.uniform_weighted_directed_lfr_graph(
N=500,
mu=0.01,
k=6,
maxk=6,
minc=10,
maxc=10,
deg_exp=1.0,
temp_dir_ID=str(np.random.randint(0, 1000)) + 'temptest',
full_path=
"/home/nathaniel/workspace/DiscreteESN/", #/N/u/njrodrig/BigRed2/topology_of_function/
weight_bounds=(-0.2, 1.0))
reservoir = np.asarray(nx.to_numpy_matrix(nx_reservoir))
esn = echostatenetwork.DESN(reservoir,
neuron_type='sigmoid',
neuron_pars={
'c': 1,
'e': 10
},
init_state=(0.0, 0.6))
esn.run_reservoir(100, record=True)
state_history = np.array(esn.network_history)
