import networkx as nx import numpy as np def icc_cascade(G, p_init): """ input: graph with prob as edge attr returns: 2D boolean matrix with indep. casc. where True means node was active at that time step p_init: proba that node in seed set """ susceptible = np.ones(G.number_of_nodes(), dtype=bool) active = np.random.rand(G.number_of_nodes()) < p_init susceptible = susceptible - active cascade = [] while sum(active) and sum(susceptible): cascade.append(active) tmp = np.zeros(G.number_of_nodes(), dtype=bool) for node in np.where(active)[0]: for edge in G.edges(node, data=True): tmp[edge[1]] += np.random.rand() < edge[2]["weight"] \ and susceptible[edge[1]] active = tmp susceptible = susceptible - active cascade.append(active) return cascade def icc_cascade_2(G, p_init): """ input: graph with prob as edge attr returns: 2D boolean matrix with indep. casc. where True means node was active at that time step p_init: proba that node in seed set """ susceptible = np.ones(G.number_of_nodes(), dtype=bool) active = np.random.rand(G.number_of_nodes()) < p_init susceptible = susceptible - active cascade = [] while active.any() and susceptible.any(): cascade.append(active) active = np.exp(np.dot(G.logmat, active)) \ < np.random.rand(G.number_of_nodes()) active = active & susceptible susceptible = susceptible - active return cascade def test(): """ unit test """ G = nx.erdos_renyi_graph(1000, 1, directed=True) G.logmat = np.zeros((G.number_of_nodes(), G.number_of_nodes())) for edge in G.edges(data=True): edge[2]['weight'] = .3*np.random.rand() G.logmat[edge[0],edge[1]] = np.log(1 - edge[2]["weight"]) import time t0 = time.time() print len(icc_cascade(G, p_init=.1)) t1 = time.time() print t1 - t0 print len(icc_cascade_2(G, p_init=.1)) print time.time() - t1 if __name__ == "__main__": test()