Since JAGS can sample from the conditional distributions of stochastic nodes,
it must have inference rules to determine these conditional distributions
(including conjugacy rules and so forth) as well as a means to store that
information in a structured manner.
I was wondering if it was possible (perhaps through some undocumented
developer feature) to print out this information. This would be useful if one
wants to try out a model in JAGS and then code an equivalent Gibbs sampler in
C, or as a means to double check one's mathematical derivation of a particular
conditional distribution in some complex hierarchical model. The
list.samplers() rjags function currently gives very basic information about
the conditional distributions of variables ("ConjugateNormal" etc.).
No. I'm sorry but there is no more detailed information on the functional form
of conditional distributions available in JAGS.
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.