R/get_gamma.R

In effectFusion: Bayesian Effect Fusion for Categorical Predictors

getGamma <- function(model) {
    
    cat <- model$categories
    
    vec_h <- rep(cat, cat - 1)/2
    vec_h_long <- rep(cat, model$diff)/2
    if (model$n_ord > 0) {
        ind_h <- cumsum(model$diff)[model$lNom - 1]
        vec_h[1:ind_h] <- 1
        vec_h_long[1:ind_h] <- 1
    }
    vec_h <- c(rep(1, model$n_cont + 1), vec_h)
    vec_h_long <- c(rep(1, model$n_cont + 1), vec_h_long)
    
    return(list(gamma = vec_h, gamma_long = vec_h_long))
    
}