BNPdensity (version 2019.9.11)

fcondYZXA: Conditional posterior distribution of the bivariate latents (Y,Z)

Description

This function simulates form the conditional posterior distribution of the latents (Y,Z).

Usage

fcondYZXA(x, distr, Tauy, Tauz, J)

Arguments

Details

For internal use.

Examples

Run this code
# NOT RUN {
## The function is currently defined as
function(x, distr = 1, Tauy, Tauz, J) {
  K <- matrix(NA, nrow = length(Tauy), ncol = length(x))
  for (i in seq(Tauy)) {
    K[i, ] <- dk(x, distr = distr, mu = Tauy[i], sigma = Tauz[i]) *
      J[i]
  }
  if (any(is.na(K))) {
    print(K, Tauy, Tauz, J)
  }
  pK <- prop.table(K, margin = 2)
  j <- apply(pK, 2, function(x) sample(length(Tauy),
      size = 1,
      prob = x
    ))
  return(matrix(c(y = Tauy[j], z = Tauz[j]),
    nrow = length(x),
    ncol = 2
  ))
}
# }

Run the code above in your browser using DataCamp Workspace