R/Gibbs_X_ij_IR.R

In CorReg: Linear Regression Based on Linear Structure Between Variables

# ' Gibbs sampler for redundant covariates
# '
Gibbs_X_ij_IR <- function(X = X, sigma = sigma, alpha = alpha, GibbsIR = FALSE) {
   mu = alpha[1] + X %*% alpha[-1]
   if (GibbsIR) {
      return(rnorm(1, mean = mu, sd = sigma))
   } else {
      return(mu)
   }
}