fcondYZXA {BNPdensity} | R Documentation |
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)
Details
For internal use.
Examples
## 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
))
}
[Package BNPdensity version 2023.3.8 Index]