VEstep: VE-step

In noisySBM: Noisy Stochastic Block Mode: Graph Inference by Multiple Testing

Description

performs one VE-step, that is, update of tau and rho

Usage

VEstep(VE, mstep, data, modelFamily, directed, fix.iter = 5)

Arguments

VE	list with variational parameters tau and rho
mstep	list with current model parameters and additional auxiliary terms
data	data vector in the undirected model, data matrix in the directed model
modelFamily	probability distribution for the edges. Possible values: Gauss, Gamma
directed	booelan to indicate whether the model is directed or undirected
fix.iter	maximal number of iterations for fixed point equation

Value

updated list VE with variational parameters tau and rho

noisySBM documentation built on Dec. 16, 2020, 5:09 p.m.