checkIsAutocorrelated: checkIsAutocorrelated - Check whether MCMC chains are...
In tmcd82070/EoAR: Evidence of Absence Regression (EoAR)

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

Run autocorrelation on every parameter in an mcmc.list and make a decision whether they are significantly auto correlated. Ideally, chains are not autocorrelated when converged.

1	checkIsAutocorrelated(obj, criterion = 0.4, lag = 1, quiet = FALSE)

`obj`	An mcmc.list object from the `rjags` package. These are output by the `coda.samples` function.
`criterion`	scalar specifying the level of acceptable autocorrelation at the assessed `lag`. If all average lag2 autocorrelations are less than this number, the whole mcmc object is deems 'random' or not autocorrelated.
`lag`	The lag at which to assess autocorrelation. This should generally be a low number like 1, 2, 5, or 10.
`quiet`	Whether to print any output on screen.

This routine calculates auto correlation separately at the specified lag for every chain in the mcmc.list, then averages autocorrelations across chains.

A list with the following components

autoCorrelated: If the average autocorrelations at the specified lag are all less than criterion, the chain is declared un-autocorrelated and is set FALSE. Otherwise, this component of the returns is TRUE.
autoCorrs: the computed autocorrelations at the specified lag. This value is the average across chains in obj.

Trent McDonald

checkIsConverged

## Not run: 
jags = jags.model(file="someJAGSFile.txt",
   data=someJAGS.data, inits=someJAGS.inits,
   n.chains=3,n.adapt=100)
update(jags, n.iter=1000)
out = coda.samples(jags,
   variable.names=c("someParameter"),
   n.iter=1000,
   thin=2)

auto <- chechIsAutocorrelated(out)

## End(Not run)