Skip to content

Latest commit

 

History

History
38 lines (28 loc) · 1.49 KB

README.md

File metadata and controls

38 lines (28 loc) · 1.49 KB
Raw

bipartite_community_detection

This repository contains R scripts for clustering biparite networks. All scripts contain a method start() with example code. Please make sure that you have installed the necessary packages.

install.packages("igraph", "dplyr", "foreach")

Nodes of the network must have an attribute 'type', indicating the mode of each node in the bipartite network (0 or 1).

 g <- read.csv("davis.csv") %>% graph_from_data_frame(directed=FALSE)
 V(g)[1:18]$type <- 1
 V(g)[1:18]$type <- 0

bipartite_cpm.R

This script implements the biclique percolation algorithm introduced by Lehman, Schwartz, and Hansen (2008).

Example: Clusters based on 4,5 bicliques clusters <- cpm(g, 4, 5)

Lehmann, S., Schwartz, M., & Hansen, L. K. (2008). Biclique communities. Physical review E, 78(1), 016108.

bipartite_modularity_optimisation.R

This script implements an adaptation of the Louvain algorithm for bipartite networks.

res <- bipartiteLouvaine(g)

subgroups <- lapply(unique(V(res)$cluster), function(c) {
   
  which(V(res)$cluster == c)
})
plot(res, mark.groups=subgroups)

  • Hecking, T., Steinert, L., Göhnert, T., & Hoppe, H. U. (2014). Incremental clustering of dynamic bipartite networks. In Proceedings of the 1st European Network Intelligence Conference (pp. 9-16). IEEE.

  • Blondel, V. D., Guillaume, J. L., Lambiotte, R., & Lefebvre, E. (2008). Fast unfolding of communities in large networks. Journal of statistical mechanics: theory and experiment, 2008(10), P10008.