jlsm (version 0.1.0)

aplsm: The Attribute Person Latent Space model

Description

Jointly model social network with multivariate attributes

Usage

aplsm(Niter, Y.i, Y.ia, D, type)

Arguments

Niter

number of iterations

Y.i

N by N matrix containing the binary social network

Y.ia

N by M matrix containing the binary multivariate attributes

D

number of dimensions in the data

type

character indicating the types of model. It could be "DD", distance by distance model, "DV", distance by vector model, "VV", vector by vector model

Value

list containing:

• lsmhEZ.i (N x D) matrix containing the posterior means of the latent person positions

• lsmhEZ.a (M x D) matrix containing the posterior means of the latent item positions

• lsmhVZ.0 (D x D) matrix containing the posterior variance of the latent person positions

• lsmhVZ.1 (D x D) matrix containing the posterior variance of the latent item positions

• lsmhAlpha.0 scaler of mean of the posterior distributions of \(\alpha.0\)

• lsmhAlpha.1 scaler of mean of the posterior distributions of \(\alpha.1\)

• lsmhKL expected log-likelihood

Examples

# NOT RUN {
attach(french)
a=aplsm(Niter=5,Y.i, Y.ia, D=2, type="DD")
# }