latent.netdep: Generate latent variable dependent network.

In netdep: Testing for Network Dependence

Description

Generate latent variable dependent network.

Usage

latent.netdep(n.node, rho = 0.3, dep.factor = 1, dep.range = c(-5, 5))

Arguments

n.node	The number of nodes in network.
rho	correlation coefficient between continuous observations and latent factor .
dep.factor	multiplicative factor applied to. If dep.factor < 0 Then A_ij ~ Bern (logistic ( dep.factor*| X_i - X_j |)) If dep.factor ≥ 0Then A_ij ~ Bern (logistic ( dep.factor / | X_i - X_j |))
dep.range	a vector specifying lower bound and upper bound for dep.factor*| X_i - X_j |. Defaults to c(-5, 5).

Value

an undirected and binary igraph object G having both Y and U as nodal attributes.

V(G)$outcome	one-dimensional continuous observations.
V(G)$latent	one-dimensional continuous latent variable dependent on V(G)$Y through rho.

Examples

library(netdep)
library(MASS)
library(mvrtn)
library(igraph)
G = latent.netdep(n.node = 100, rho = 0.5, dep.factor = 1)

netdep documentation built on May 2, 2019, 5:52 a.m.