loe: spec.emb – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

spec.emb

Spectral embedding with the normalized or the unnormalized Laplacian (Laplacian eigenmaps).

Performs spectral embedding for a given adjacency matrix.

spec.emb(A, p, norm = TRUE)

`A`
`p`	The number of dimensions.
`norm`	If `TRUE`, then the normalized Laplacian is used for embedding. If `FASLE`, then the unnormalized Laplacian is used for embedding.

The corrdinate matrix with p columns whose rows give the coordinates of the vertexes.

Yoshikazu Terada

library(igraph)
ADM <- as.matrix( get.adjacency(graph.famous("Icosahedral")) )

#Apply some graph embedding methods
LE <-spec.emb(A=ADM,2,norm=FALSE)

Local Ordinal Embedding

v1.1

GPL (>= 2)

Authors

Yoshikazu Terada, Ulrike von Luxburg

Initial release

2016-02-09