Awesome

ngraph.coarsen

Given a community structure creates a coarse graph

usage

// Let's say you have detected community of a graph using
// ngraph.cw or ngraph.louvain:
var community = detectCommunities(srcGraph);

// To build a coarse graph:
var coarsen = require('ngraph.coarsen');
var coarseGraph = coarsen(srcGraph, community)

Each node in the coarse graph is a community inside srcGraph. Link between communities exists if members of each community are connected in the srcGraph.

• node.data property is a Set of node ids from the srcGraph;
• link.data is a number, that shows total number of connections (or weights) between members of communities.

Note: each node in the coarse graph can also have a link with itself (a self-reference). The weight of this link is equal to weight of all weights within community.

Array of subgraphs

You can also get plain array of subgraphs for each found community:

var coarseGraph = coarsen(srcGraph, community)
var subgraphs = coarsen.getSubgraphs(coarseGraph);
// subgraphs is array, where each element has:
// * id - community id
// * graph - graph with nodes and edges that represent a subgraph of srcGraph

See also

• https://github.com/anvaka/ngraph.graph - graph structure
• https://github.com/anvaka/ngraph.cw - label-propagation based community detection
• https://github.com/anvaka/ngraph.louvain - Louvain method for community detection (based on modularity optimization)

license

MIT