R/momentsToDistArgs.R

In markwh/conditionalSample: Allows samples to be drawn from (generalized) linear models

#' Derive arguments to distribution functions from distribution moments
#'
#' Returns a list of distribution parameters for R distribution calls.
#'
#' @param distr character. A distribution R recognizes (currently only supports
#' 'norm', 'gamma', 'poisson')
#' @param moments a numeric vector, matrix, or data.frame with mean as column 1,
#' variance as column 2, etc. Currently only first 2 moments are used.
#' @return list of distribution parameters for R param calls.
#' @export

momentsToDistArgs = function(distr, moments) {
    if (is.vector(moments)) 
        moments = matrix(moments, nrow = 1)
    moments = as.matrix(moments)
    switch(distr, norm = list(mean = moments[, 1], sd = sqrt(moments[, 2])), 
           gamma = list(shape = moments[, 1]^2/moments[, 2], scale = moments[, 2]/moments[, 1]), 
           pois = list(lambda = moments[, 1]))
}


# foo = momentsToDistArgs('norm', c(0, 4)) bar = momentsToDistArgs('gamma', data.frame(means = runif(10, 1, 10), vars = runif(10, 10, 20)))
# do.call('rnorm', c(list(n = 10), foo)) do.call('rgamma', c(list(n = 10), bar))