fit.wrap: Internal function

In rauschenberger/semisup: Semi-Supervised Mixture Model

Description

This function fits the semi-supervised mixture model multiple times. It is called by mixtura and scrutor.

Usage

fit.wrap(y, z, dist, phi, pi, gamma, starts = 1, it.em = 100, epsilon = 1e-04)

Arguments

y	observations: numeric vector of length n
z	class labels: integer vector of length n, with entries 0, 1 and NA
dist	distributional assumption: character "norm" (Gaussian), "nbinom" (negative bionomial), or "zinb" (zero-inflated negative binomial)
phi	dispersion parameters: numeric vector of length q, or NULL
pi	zero-inflation parameter(s): numeric vector of length q, or NULL
gamma	offset: numeric vector of length n, or NULL
starts	restarts of the EM algorithm: positive integer (defaults to 1)
it.em	(maximum) number of iterations in the EM algorithm: positive integer (defaults to 100)
epsilon	convergence criterion for the EM algorithm: non-negative numeric (defaults to 1e-04)

Details

The distributions are parametrised as follows:

• Gaussian
  y ~ N(mean,sd^2)
E[y]=mean
Var[y]=sd^2

• Negative binomial
  y ~ NB(mu,phi)
E[y]=mu
Var[y]=mu+phi*mu^2

• Zero-inflated negative binomial
  y ~ ZINB(mu,phi,pi)
E[y]=(1-pi)*mu

Value

This function returns the parameter estimates, the posterior probabilities, and the likelihood.

posterior	probability of belonging to class 1: numeric vector of length n
converge	path of the log-likelihood: numeric vector with maximum length it.em
estim0	parameter estimates under H0: data frame
estim1	parameter estimates under H1: data frame
loglik0	log-likelihood under H0: numeric
loglik1	log-likelihood under H1: numeric
lrts	likelihood-ratio test statistic: positive numeric

See Also

This is an internal function. The user functions are mixtura and scrutor.

Examples

# data simulation
n <- 100
z <- rep(0:1,each=n/2)
y <- rnorm(n=n,mean=2*z,sd=1)
z[(n/4):n] <- NA

# model fitting
fit.wrap(y,z,dist="norm")

rauschenberger/semisup documentation built on May 11, 2020, 4 p.m.