graphclust (version 1.3)

fitSimpleSBM: Fit a stochastic block model to every network in a collection of networks.

Description

Applies the variational EM-algorithm implemented in the package blockmodels to every network.

Usage

fitSimpleSBM(
  allAdj,
  directed = TRUE,
  nbSBMBlocks = Inf,
  nbCores = 1,
  outCountStat = TRUE
)

Value

list of count statistics for every network or list of parameters of the stochastic block models fitted to every network.

Arguments

allAdj

list of adjacency matrices

directed

Networks are directed (TRUE by default) or undirected (FALSE).

nbSBMBlocks

upper bound for the number of blocks in the SBMs of the mixture components. Default is Inf

nbCores

number of cores for parallelization.

outCountStat

If TRUE (default), the output is a list of count statistics for every network. If FALSE, the output is a list of parameters of the stochastic block models fitted to every network.

Examples

theta <- list(pi=c(.5,.5), gamma=matrix((1:4)/8,2,2))
obs <- rCollectSBM(rep(10,4), theta)$listGraphs
res <- fitSimpleSBM(obs, outCountStat=FALSE, nbCores=2)