Introduction to the Modeling and Analysis of Complex Systems

16.4. SIMULATING ADAPTIVE NETWORKS 367you are interested in more details of this, you should read Zachary’s original paper [59],which contains some more interesting stories.This data looks perfect for our modeling purpose. We can set up the initialize functionusing this ’club’ property to assign binary states to the nodes. The implementationof dynamical equations are straightforward; we just need to discretize time and simulatethem just as a discrete-time model, as we did before. Here is a completed code:Code 16.17: net-diffusion-adaptive.pyimport matplotlibmatplotlib.use(’TkAgg’)from pylab import *import networkx as nxdef initialize():global g, nextgg = nx.karate_club_graph()for i, j in g.edges_iter():g.edge[i][j][’weight’] = 0.5g.pos = nx.spring_layout(g)for i in g.nodes_iter():g.node[i][’state’] = 1 if g.node[i][’club’] == ’Mr. Hi’ else 0nextg = g.copy()def observe():global g, nextgcla()nx.draw(g, cmap = cm.binary, vmin = 0, vmax = 1,node_color = [g.node[i][’state’] for i in g.nodes_iter()],edge_cmap = cm.binary, edge_vmin = 0, edge_vmax = 1,edge_color = [g.edge[i][j][’weight’] for i, j in g.edges_iter()],pos = g.pos)alpha = 1 # diffusion constantbeta = 3 # rate of adaptive edge weight changegamma = 3 # pickiness of nodesDt = 0.01 # Delta t