R/simcubshe.R

In CUB: A Class of Mixture Models for Ordinal Data

#' @title Simulation routine for CUB models with shelter effect
#' @aliases simcubshe
#' @description Generate \eqn{n} pseudo-random observations following the given CUB distribution
#'  with shelter effect.
#' @keywords distribution
#' @usage simcubshe(n,m,pai,csi,delta,shelter)
#' @export simcubshe
#' @import stats
#' @param n Number of simulated observations
#' @param m Number of ordinal categories
#' @param pai Uncertainty parameter
#' @param csi Feeling parameter
#' @param delta Shelter parameter
#' @param shelter Category corresponding to the shelter choice
#' @seealso \code{\link{probcubshe1}}, \code{\link{probcubshe2}}, \code{\link{probcubshe3}}
#' @examples
#' n<-300
#' m<-9
#' pai<-0.7
#' csi<-0.3
#' delta<-0.2
#' shelter<-3
#' simulation<-simcubshe(n,m,pai,csi,delta,shelter)
#' plot(table(simulation),xlab="Ordinal categories",ylab="Frequencies")


simcubshe <-
function(n,m,pai,csi,delta,shelter){
  dicopai<-runif(n)<pai
  dicodelta<-runif(n)<delta
  cub00<-dicopai*(1+rbinom(n,m-1,1-csi))+(1-dicopai)*sample(m,n,replace=TRUE)
  ord<-(1-dicodelta)*cub00+dicodelta*shelter
  return(ord)
}