Adding a new prior_samples MCMC sampler, which allows defining the prior distribution of a model node using a set of numeric values. Values are provided either as a vector (in the scalar case), or as a matrix (in the multidimensional case).
Draws from this numeric structure are taken either sequentially (the default), or drawn randomly, as controlled by the randomDraws control-list argument.
prior_samples samplers are (by default) reordered to operate first among MCMC samplers, however this is controlled by a new system option MCMCorderPriorSamplesSamplersFirst.
prior_samples samplers are able to operate on non-stochastic nodes, in particular on RHSonly nodes.
Adding a new
prior_samplesMCMC sampler, which allows defining the prior distribution of a model node using a set of numeric values. Values are provided either as a vector (in the scalar case), or as a matrix (in the multidimensional case).Draws from this numeric structure are taken either sequentially (the default), or drawn randomly, as controlled by the
randomDrawscontrol-list argument.prior_samplessamplers are (by default) reordered to operate first among MCMC samplers, however this is controlled by a new system optionMCMCorderPriorSamplesSamplersFirst.prior_samplessamplers are able to operate on non-stochastic nodes, in particular onRHSonlynodes.Documentation and testing added.
See roxygen in
MCMC_samplers.Rfor full details.