R/tef_penalizedErr.R

In akcochrane/TEfits: Time-evolving model fits

#' Penalize error for being close to a boundary
#'
#' \code{\link{TEfit}} internal.
#' Formally, `pErr=Err/dbeta(par,dbeta_penalty,dbeta_penalty)` when Err is positive, where
#' par is the (often rate) parameter normalized between 0 [lower bound] and 1 [upper bound]
#'
#' @param boundedPar Value of the parameter that has bounds.
#' @param errPar Value of the error to be penalized.
#' @param loBound Lower parameter boundary.
#' @param upBound Upper parameter boundary.
#' @param dbeta_penalty The multiplicative error penalty at approximately 5.3 percent away from the boundary. Error increases with increasing proximity to a bound.
#'
#' @export
#'
tef_penalizedErr <-  function(boundedPar,errPar,loBound,upBound,dbeta_penalty=1.001){
  boundedPar <- (boundedPar-loBound)/(upBound-loBound)
  if(sign(errPar)==1){
  p_err <- errPar/dbeta(boundedPar,dbeta_penalty,dbeta_penalty)
  }else{
  p_err <- errPar*dbeta(boundedPar,dbeta_penalty,dbeta_penalty)
  }
  return(p_err)
}