calculateDIC: Compute DIC for fitted mixture model

Description

Computes and returns the Deviance Information Critereon (DIC) as suggested by Celeaux et al (2006) as their DIC$_4$ for Bayesian mixture models

Usage

calculateDIC(mcmc.mixture, model, priors, seg.ratios, chain=1, print.DIC=FALSE)

Arguments

mcmc.mixture

Object of type segratioMCMC produced by coda usually by using readJags

model

object of class modelSegratioMM specifying model parameters, ploidy etc

priors

Object of class priorsSegratioMM

seg.ratios

Object of class segRatio contains the segregation ratios for dominant markers and other information such as the number of dominant markers per individual

chain

Which chain to use when compute dosages (Default: 1)

print.DIC

Whether to print DIC

Value

A scalar DIC is returned

References

G Celeaux et. al. (2006) Deviance Information Criteria for Missing Data Models Bayesian Analysis 4 23pp
D Spiegelhalter et. el. (2002) Bayesian measures of model complexity and fit JRSS B 64 583--640

Examples

Run this code

# NOT RUN {
## simulate small autooctaploid data set
a1 <- sim.autoMarkers(8,c(0.7,0.2,0.1),n.markers=100,n.individuals=50)

## compute segregation ratios
sr <-  segregationRatios(a1$markers)

## set up model, priors, inits etc and write files for JAGS
x <- setModel(3,8)
x2 <- setPriors(x)
dumpData(sr, x)
inits <- setInits(x,x2)
dumpInits(inits)
writeJagsFile(x, x2, stem="test")

# }
# NOT RUN {
## run JAGS
small <- setControl(x, burn.in=200, sample=500)
writeControlFile(small)
rj <- runJags(small)  ## just run it
print(rj)

## read mcmc chains and print DIC
xj <- readJags(rj)
print(calculateDIC(xj, x, x2, sr))
# }