dMCMCplot: Create a multiple diagnostic plot
In MarcoDVisser/dMCMC: A compilation of diagnostic tools for mcmc evaluation

Description Usage Arguments Details Value Author(s) See Also Examples

dMCMCplot Create the untimate diagnostic plot for a MCMC chain

1 2	dMCMCplot(mcmcobject = NULL, interest = NULL, model = NULL, Rprior = NULL, ...)

`mcmcobject`	a mcmc object (list of samples).
`interest`	a set of parameters of interest. These should have true priors (e.g. can be directly related to a prior distribution). Lower-level priors, that depend on a hyperprior, will fail as dMCMC does not attempt to understand the hierachical structure of the model file.
`model`	the path to the BUGs/JAGS model text file with the priors specified. This model must be the one that generated the MCMC list.
`Rprior`	the prior function directly specified by the user (as a standard R probabilty density function e.g. prior=dunif).
`...`	additional parameters (unused). diagnostic plots on. Names must be equal to the names in the jags/bugs model. If length > 1, one plot is produced per parameter of interest, in this case output to a pdf is advised. Otherwise the function forces par(ask=TRUE).

Creates a multipanel graphic with diagnostics on parameter samples

A multipanel plot

Marco D. Visser

jags.model

sink("examp.txt")
cat( "
model {
	for (i in 1:N) {
		x[i] ~ dnorm(mu, tau)
	}
	mu ~ dnorm(0, .0001)
	tau <- pow(sigma, -2)
	sigma ~ dunif(0, 100)
    } ",fill=TRUE)
     sink()

jags <- jags.model('examp.txt',
                  data = list('x' = rnorm(100,2,2),
                              'N'=100),
                  n.chains = 4,
                  n.adapt = 100)

mysamples <- coda.samples(jags, c('mu', 'tau'),100)

dMCMCplot(mysamples,"mu","examp.txt")