simulate {missSBM} | R Documentation |
Generates a realization (blocks and adjacency matrix) of a Stochastic Block model
simulate(nNodes, mixtureParam, connectParam, directed = FALSE, covariates = NULL, covarParam = NULL)
nNodes |
The number of nodes |
mixtureParam |
The mixture parameters |
connectParam |
The connectivity matrix (inter/intra clusters probabilities. provided on a logit scale for a model with covariates) |
directed |
Boolean variable to indicate whether the network is directed or not. Default to |
covariates |
A list with M entries (the M covariates). Each entry of the list must be an N x N matrix. |
covarParam |
An optional vector of parameters associated with the covariates, with size M |
an object with class SBM_sampler
The class SBM_sampler
## SBM parameters directed <- FALSE N <- 300 # number of nodes Q <- 3 # number of clusters M <- 2 # two Gaussian covariates alpha <- rep(1, Q)/Q # mixture parameters pi <- diag(.45, Q) + .05 # connectivity matrix eta <- rnorm(M, -1, 1) # covariate parametes gamma <- log(pi/(1-pi)) # logit transform of pi for the model with covariates X <- replicate(M, matrix(rnorm(N * N ,mean = 0, sd = 1), N, N), simplify = FALSE) ## draw a SBM without covariates sbm <- missSBM::simulate(N, alpha, pi, directed) ## draw a SBM model with node-centred covariates sbm_cov <- missSBM::simulate(N, alpha, gamma, directed, X, eta) old_param <- par(mfrow = c(1,2)) plot(sbm) plot(sbm_cov) par(old_param)