R/parms_from_priors.R

In carolyner/imabc: Incremental Mixture Approximate Bayesian Computation (IMABC)

parms_from_priors <- function(parm_df, name_parms, prior_list, sampling) {
  # Generate parameter values based on a quantile function
  if (is.null(sampling)) {
    # Generate values based on sample size desired (should only be used on fixed parameters)
    parm_df[1:nrow(parm_df), (name_parms) := lapply(name_parms, FUN = function(x, prior_funcs, n) {
      prior_funcs[[x]]$quantile_function(n)

    }, prior_funcs = prior_list, n = nrow(parm_df))]
  } else {
    # Generate values based on sampling distribution
    parm_df[1:nrow(sampling), (name_parms) := lapply(name_parms, FUN = function(x, prior_funcs, df) {
      prior_funcs[[x]]$quantile_function(df[, which(name_parms %in% x)])

    }, prior_funcs = prior_list, df = sampling)]
  }

  return(parm_df)
}