R/samplingPosterior.R

In PratheepaJ/BARBI: A Bayesian Approach to Contamination Removal in Molecular Microbial Studies

#' Sampling posterior
#'
#' Sampling from the marginal posterior for the true intensity
#'
#' @param gammaPrior_Cont A list of estimated distribution parameters for the
#'        contamination intensity in thespecimen samples in a particular
#'        block.
#' @param itera Numeric, number of MCMC samples.
#' @inheritParams alphaBetaContInPlasma
#' @inheritParams MH_MCMC
#' @importFrom parallel mclapply
#' @return A list of marginal posterior for the true intensity
#' @export
samplingPosterior = function(psPlByBlock,
                             blk,
                             gammaPrior_Cont,
                             itera = 10000) {

    sampleLst = seq(1, nsamples(psPlByBlock[[blk]]))
    sampleLst = as.list(sampleLst)

    sub_sampling_pos = function(x, gammaPrior_Cont, blk, itera) {
        sam = x
        taxa_list = as.list(1:ntaxa(psPlByBlock[[blk]]))

        taxa_post = lapply(taxa_list, function(taxa) {
            chain = MH_MCMC(itera = itera,
                            k = as.numeric(gammaPrior_Cont[[sam]]$kij[taxa]),
                            al_c = gammaPrior_Cont[[sam]]$alpha_ij_c[taxa],
                            be_c = gammaPrior_Cont[[sam]]$beta_ij_c[taxa],
                            startvalue_lamda_r = 0)
            return(chain)
        })

        return(taxa_post)
    }

    taxa_post_all_sam = lapply(sampleLst,
                               FUN = sub_sampling_pos,
                               gammaPrior_Cont = gammaPrior_Cont,
                               blk = blk,
                               itera = itera)

    return(taxa_post_all_sam)
}