R/initialize_values.R

In philpatton/cowbird: Recreate analysis in Patton et al (TBD)

#' Initialize values for MCMC
#'
#' Initializes values for MCMC, namely, the occurrence states of the parasites
#' and hosts.
#'
#' @param data_list list made from \code{make_data_list()}, containing detection
#' matrices for cowbird and hosts
#' @param fix_seed logical. should the seed be set for reproducible results?
#'
#' @return list, suitable for use in \code{rjags::jags.model()}
#' @export

initialize_values <- function(data_list, fix_seed = FALSE) {

    X <- data_list$X
    Y <- data_list$Y

    # output from calc_naive_state_() functions is logical
    naive_state_host <- calc_naive_state_host(X) * 1
    naive_state_cow <- calc_naive_state_cowbird(Y) * 1

    inits <-  list(z = naive_state_cow, tau = naive_state_host)

    if (fix_seed) {
        seed_params <- list (
            .RNG.name = "base::Marsaglia-Multicarry",
            .RNG.seed = 17
        )

        inits <- c(inits, seed_params)

    }

    inits

}