CODEWITHSUNDEEP
rigraph
Michael Davidson
Michael Davidson

Reputation: 1411

Subset igraph graph by label

I am trying to subset a igraph graph by an edge characteristics (like its label). In the reproducible example I have shamelessly stolen from another post with a little modification, I would like to be able to separate the Best Friend ties (BF) from the Family ties (FAM):

edges <- matrix(c(103, 86, 24, 103, 103, 2, 92, 103, 87, 103, 103, 101, 103, 44), ncol=2, byrow=T)
g <- graph(as.vector(t(edges)))
E(g)[c(2:4,7)]$label<-"FAM"
E(g)[c(1,5,6)]$label<-"BF"

The best I can do so far is display the edges which have one type of tie:

E(g)[E(g)$label=="BF"]
V(g)[E(g)$label=="BF"]

Upvotes: 17

Views: 10200

Answers (2)

Solomon
Solomon

Reputation: 956

how about:

gfam <- subgraph.edges(graph=g, eids=which(E(g)$label=="FAM"), delete.vertices = TRUE)
gbf <- subgraph.edges(graph=g, eids=which(E(g)$label=="BF"), delete.vertices = TRUE)

Suggestion for igraph/network analysis tutorial/shameless plug:

https://solomonmg.github.io/post/working-with-bipartite-affiliation-network-data-in-r/

Stanford Network Analysis in R lab: https://www.stat.cmu.edu/~brian/780/stanford%20social%20network%20labs/

Upvotes: 25

Gabor Csardi
Gabor Csardi

Reputation: 10825

I suggest you read ?V and ?E to see how to select edges and vertices. A quite compact and readable solution to your question is 

subgraph.edges(g, E(g)[label=="FAM"])
subgraph.edges(g, E(g)[label=="BF"])

This removes the vertices as well, if they don't have an incident edge of the specified label. See ?subgraph.edges for details.

Upvotes: 19

Related Questions