latentnet (version 0.6-0)

latentcluster: Latent cluster models for networks

Description

latentcluster() is a term to the function ergm to fit a latent postion model to a given network, $g$. ergm returns a Bayesian model fit based on a Monte Carlo scheme The default prior specifications are diffuse. An approximate MLE fit is also returned.

The ergm specifies models via: g ~ latentcluster() where g is a network object For the list of possible , see below.

Usage

latentcluster(k=2, ngroups, z.prior.mu=0, z.prior.sd=10, b.delta=0.5,
               b.prior.mu=0, b.prior.sd=10,
               Sigprior = qchisq(0.05,3),
               muSigprior = 2, dirprior=3,
               alphaprior=3,
               chisqprop = 6, thetaprop=0.1, ...)

Arguments

k

Dimension of the latent space.

ngroups

Number of clusters in the latent space.

z.prior.mu

Prior mean for each dimension of the latent positions. If a constant is passed it is used for each dimension.

z.prior.sd

Prior standard deviation for each dimension of the latent positions. If a constant is passed it is used for each dimension.

b.delta

Standard deviation of the deviance for covariate parameters. If a constant is passed it is used for each dimension.

b.prior.mu

Prior mean for the covariate parameters. If a constant is passed it is used for each dimension.

b.prior.sd

Prior standard deviation for the covariate parameters. If a constant is passed it is used for each dimension.

Sigprior

Prior standard deviations for the node positions relative to the cluster mean. If a constant is passed it is used for each dimension.

muSigprior

Prior standard deviations for the node positions relative to the cluster mean. If a constant is passed it is used for each dimension.

dirprior

Prior standard deviations for the node positions relative to the cluster mean. If a constant is passed it is used for each dimension.

alphaprior

Prior standard deviations for the node positions relative to the cluster mean. If a constant is passed it is used for each dimension.

chisqprop

Prior standard deviations for the node positions relative to the cluster mean. If a constant is passed it is used for each dimension.

thetaprop

Prior standard deviations for the node positions relative to the cluster mean. If a constant is passed it is used for each dimension.

...

Specific to the model term.

Value

• ergm returns an object of class 'ergm' that is a list.

References

Peter D. Hoff, Adrian E. Raftery and Mark S. Handcock. Latent space approaches to social network analysis. Journal of the American Statistical Association, Dec 2002, Vol.97, Iss. 460; pg. 1090-1098.

Mark S. Handcock, Adrian E. Raftery and Jeremy Tantrum. Model-Based Clustering for Social Networks. Working Paper Number 46, Center for Statistics and the Social Sciences, University of Washington, April 2005.

See Also

latent, plot.ergm, sna, network, terms.ergm

Examples

#
# Using Sampson's Monk data, lets fit a 
# simple latent position model
#
data(sampson)
#
# Get the group labels
samp.labs <- substr(get.vertex.attribute(samplike,"group"),1,1)
#
# Using Sampson's Monk data, lets fit a latent clustering model
#
samp.fit <- ergm(samplike ~ latentcluster(k=2, ngroups=3), burnin=10000,
                 MCMCsamplesize=2000, interval=30)
#
# See if we have convergence in the MCMC
mcmc.diagnostics(samp.fit)
#
# Lets look at the goodness of fit:
#
plot(samp.fit,label=samp.labs, vertex.col="group")
plot(samp.fit,pie=TRUE,label=samp.labs)
plot(samp.fit,density=c(2,2))
plot(samp.fit,contours=5,contour.color="red")
plot(samp.fit,density=TRUE,drawarrows=TRUE)
add.contours(samp.fit,nlevels=8,lwd=2)
points(samp.fit$Z.mkl,pch=19,col=samp.fit$class)