Just came across this very easy library for community detection https://sites.google.com/site/findcommunities/ https://bitbucket.org/taynaud/python-louvain/src. Here’s how to create a graph, detect communities in it, and then visualize with nodes colored by their community in less than 10 lines of python:

import networkx as nx
import community

G = nx.random_graphs.powerlaw_cluster_graph(300, 1, .4)

part = community.best_partition(G)
values = [part.get(node) for node in G.nodes()]

nx.draw_spring(G, cmap = plt.get_cmap('jet'), node_color = values, node_size=30, with_labels=False)

community structuree

It’s easy to get modularity to:

mod = community.modularity(part,G)
print("modularity:", mod)

gave modularity: 0.8700238252368541.

blog comments powered by Disqus


16 June 2014