[R-sig-ME] bayes factors for MCMCglmm
j.hadfield at ed.ac.uk
Tue Aug 28 09:51:29 CEST 2012
The fact that the deviance is not calculated after marginalising the
random effects may also be an issue for Bayes factors, as it can be
with DIC. But not sure.
Quoting Ben Bolker <bbolker at gmail.com> on Tue, 28 Aug 2012 02:43:23
> Mikhail Matz <matz at ...> writes:
>> Hello - is it possible to calculate hayes factors for MCMCglmm
>> objects, to compare alternative models? Mikhail
> Well, MCMCglmm gives you a posterior sample for the deviance, so
> in principle you could calculate the harmonic mean of the likelihood,
> model1<-MCMCglmm(PO~1, random=~FSfamily, data=PlodiaPO, verbose=FALSE)
> likchain <- exp(-model1$Deviance/2)
> these are ridiculously big numbers -- that may have something
> to do with the way the deviance is scaled.
> However, harmonic means are known to be really, really
> dodgy estimates of the Bayes factor -- see e.g.
> (warning, broken URL)
> Perhaps someone else will contribute something more useful.
> You could also use DIC to compare models, although it has
> its own issues ...
> R-sig-mixed-models at r-project.org mailing list
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
More information about the R-sig-mixed-models