Community detection algorithms

[tecosaur]

There seems to be a dearth of community detection algorithms in the current Graphs.jl. Hopefully this can be remedied in the future, and to that end I thought I'd list the ones that NetworkX currently has for comparison:

• [ ] Kernighan–Lin bipartition
• [x] K-Clique (already in Graphs.jl as clique percolation)
• [ ] Greedy modular communities
• [ ] Lukes exact optimal weighted tree partitioning
• [x] Label propagation (already in Graphs.jl)
• [ ] Leiden Community Detection
• [ ] Fluid Communities
• [ ] Girvan–Newman (@leerosenthalj still up for this?)
• [x] Spectral clustering (in https://github.com/JuliaGraphs/CommunityDetection.jl)

There are also some in iGraphs that I think would be worth having:

• [ ] Louvain Community Detection
• [ ] Walktrap (short random walks)
• [ ] Spin-glass model and simulated annealing
• [ ] Optimal modularity
• [ ] Edge betweenness
• [ ] Eigenvector of the modularity matrix

[gdalle]

@olegfafurin let's start with greedy modularity optimization?

[leerosenthalj]

I am still interested in implementing Girvan-Newman! I need a week before I can make steady progress, but would be happy to take a crack at it.

[gdalle]

Sounds good!