Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

re-computation of totalGraphWeight LouvainCore.louvain #87

Open
5abhim opened this issue Jul 24, 2015 · 2 comments
Open

re-computation of totalGraphWeight LouvainCore.louvain #87

5abhim opened this issue Jul 24, 2015 · 2 comments

Comments

@5abhim
Copy link

5abhim commented Jul 24, 2015

In LouvainCore.louvain, we compute totalGraphWeight at very first step which is nothing but 2 * sum-of-all-edges-weights. Why do we need to compute this again and again whenever we call LouvainCore.louvain from outside do-while loop. I think this graph weight is fixed because we never delete any edges nor add any edge, we just compress node (means we convert edges into self edge) which should not change totalGraphWeight?
Please let me know, if I am thinking in wrong direction.

@5abhim
Copy link
Author

5abhim commented Jul 27, 2015

Hello Eric and others,
It would be great if you can take some time out and look at my understanding of how total-graph-weight is computed.
Thanks in advance.

@eric-kimbrel
Copy link
Contributor

In our implementation we are counting only links between nodes, and ignoring self loops. Therefore we must count the weight of these links each time the graph is compressed to change the network topology. This is my understanding of the original algorithm from the inspiring paper:

"Fast unfolding of communities in large networks Vincent D. Blondel, Jean-Loup Guillaume, Renaud Lambiotte, Etienne Lefebvre"

Its likely that our implementation is not optimal and could be imporoved , but we have had good results with this method.

Perhaps we could use a variable name like "totalLinkWeight" instead to be more clear.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants