Bipartite example
-----------------

:download:`Downloadable Source Code <bipartite/bipartite.py>` 

.. image:: bipartite/bipartite.png
    :width: 80 %
    
::

    import EoN
    import networkx as nx
    import random
    import matplotlib.pyplot as plt
    
    
    G= nx.bipartite.configuration_model([1,11]*2000, [3]*8000)
    #the graph now consists of two parts.  The first part has 2000 degree 1 nodes
    #and 2000 degree 11 nodes.  The second has 8000 degree 3 nodes.
    #there are 24000 edges in the network.
    #
    #  We assume the first ones are twice as infectious as the second ones.
    #
    
    for node in G:
        if G.degree(node) in [1,11]:
            G.node[node]['type'] = 'A'
        else:
            G.node[node]['type'] = 'B'
            
    #We have defined the two types of nodes.
    
    #now define the transmission and recovery functions:
    def trans_time_function(source, target, tau):
        if G.node[source]['type'] is 'A':
            return random.expovariate(2*tau)
        else:
            return random.expovariate(tau)
        
    def rec_time_function(node, gamma):
        return random.expovariate(gamma)
    
    tau = 0.4
    gamma = 1.
    sim = EoN.fast_nonMarkov_SIR(G, trans_time_function, rec_time_function, 
                                trans_time_args=(tau,), rec_time_args=(gamma,),
                                rho = 0.01, return_full_data=True)
                                
    t, S, I, R = sim.summary()
    plt.plot(t, I, label='Total Infecteds')
    
    t1, S1, I1, R1 = sim.summary(nodelist = [node for node in G if G.node[node]['type']=='A'])
    plt.plot(t1, I1, label = 'Partition 1')
    
    t2, S2, I2, R2 = sim.summary(nodelist = [node for node in G if G.node[node]['type']=='B'])
    plt.plot(t2, I2, label = 'Partition 2')
    
    plt.legend()
    plt.xlabel('$t$')
    plt.ylabel('Infecteds')
    plt.savefig('bipartite.png')