e.step {mixture}R Documentation

E-Step

Description

Carries out the E-step for EM algorithm

Usage

e.step(data=NULL, gpar=NULL, labels=NULL, v=1)

Arguments

data

A matrix or data frame such that rows correspond to observations and columns correspond to variables. Note that this function currently only works with multivariate data p > 1.

gpar

A list of the model parameters.

labels

A vector of groups labels. If NULL none are known.

v

The value for deterministic annealing. If v=1 the standard estimate is used.

Details

Carries out the E-step for EM algorithm

Value

A nxG matrix of weights.

Examples

data("x2")
u0 = runif(nrow(x2))
m0 = m.step(data=x2, covtype="VVV", w=cbind(u0,1-u0), D=NULL, mtol=1e-8, mmax=10)
w0 = e.step(data=x2, gpar=m0, labels=NULL, v=1)

[Package mixture version 1.5 Index]