Girvan-Newman record of removed edges

[Daniel Burns]

Hello,

I am using the Girvan-Newman algorithm to break up a network into communities.

I am using the default edge_betweeness_centrality as the method of selecting edges to remove in order to find the communities and using edge weights in the calculation.

Once I have the number of communities I want, is there a way to see which edges were removed and in which order?  

Thank you,

Dan

[Dan Schult]

I think that is precisely the kind of personalization that was intended for the argument: most_valuable_edge

The default value calls networkx.edge_betweenness_centrality() and returns the edge with the 

largest centrality. So, you could write a function to do exactly that same calculation, but to also store the

edge in a list accessible to the function, but also to you outside of that function.

edges_removed = []

def recorded_most_valuable_edge(G)

    betweenness = nx.edge_betweenness_centrality(G) 

    max_edge = max(betweenness, key=betweenness.get)

    edges_removed.append(max_edge)

    return max_edge

comp = girvan_newman(G, most_valuable_edge=recorded_most_valuable_edge)

print(edges_removed)

[Daniel Burns]

Thank you!  Works great - I added the list as an argument to the function before trying it verbatim.

Thank you,

Dan