Builds an igraph graph from a co-occurrence data.table (edge list with
from, to, weight columns) and applies a community detection
algorithm. Supported methods: infomap, label propagation, walktrap,
and spinglass.
Usage
detect_skill_communities(cooc, method = "infomap")
Arguments
- cooc
A data.table with columns from, to, and weight
representing the co-occurrence edge list. Weights must be positive.
- method
Character, community detection method. One of "infomap",
"label_prop", "walktrap", "spinglass". Default: "infomap".
Value
A named list with:
- graph
An igraph graph object built from the edge list.
- communities
The community detection result object (class
depends on the method used).
- membership
A named integer vector mapping each vertex to its
community ID.