R/GIBmix.R

In IBclust: Information Bottleneck Methods for Clustering Mixed-Type Data

GIBmix <- function(X, ncl, beta, alpha, catcols, contcols, randinit = NULL,
                   lambda = -1, s = -1, scale = TRUE,
                   maxiter = 100, nstart = 100,
                   verbose = FALSE) {
  
  # Validate inputs
  if (!is.data.frame(X)) {
    stop("Input 'X' must be a data frame.")
  }
  
  if (!is.numeric(ncl) || ncl <= 1 || ncl != round(ncl)) {
    stop("Input 'ncl' must be a positive integer greater than 1.")
  }
  
  if (!is.numeric(beta) || beta <= 0) {
    stop("Input 'beta' must be a positive number.")
  }
  
  if (!is.numeric(alpha) || alpha < 0) {
    stop("Input 'alpha' must be a non-negative number.")
  }
  
  if (!all(catcols %in% seq_along(X))) {
    stop("Some 'catcols' indices are out of bounds or invalid.")
  }
  
  if (!all(contcols %in% seq_along(X))) {
    stop("Some 'contcols' indices are out of bounds or invalid.")
  }
  
  if (any(duplicated(c(catcols, contcols)))) {
    stop("'catcols' and 'contcols' must not overlap.")
  }
  
  if (!is.logical(scale)) {
    stop("'scale' must be a logical value (TRUE or FALSE).")
  }
  
  if (!is.numeric(maxiter) || maxiter <= 0 || maxiter != round(maxiter)) {
    stop("'maxiter' must be a positive integer.")
  }
  
  if (!is.numeric(nstart) || nstart <= 0 || nstart != round(nstart)) {
    stop("'nstart' must be a positive integer.")
  }
  
  if (!is.null(randinit) && (!is.numeric(randinit) || length(randinit) != nrow(X))) {
    stop("'randinit' must be a numeric vector with length equal to the number of rows in 'X', or NULL.")
  }
  
  # Validate lambda
  if (!is.numeric(lambda) ||
      !(length(lambda) == 1 || length(lambda) == length(catcols)) ||
      any(lambda <= 0 & lambda != -1)) {
    stop("'lambda' must be either a single numeric value (-1 for automatic selection or a positive value) or a numeric vector with positive values matching the number of 'catcols'.")
  }
  
  # Additional check for maximum lambda value for nominal variables
  if (length(lambda) > 1 && length(lambda) == length(catcols)) {
    max_lambda <- sapply(catcols, function(col) {
      l <- length(unique(X[, col]))
      (l - 1) / l
    })
    if (any(lambda > max_lambda)) {
      stop("'lambda' values for nominal variables must not exceed their maximum allowable value of (l - 1)/l, where l is the number of categories in the variable.")
    }
  }
  
  
  # Validate s
  if (!is.numeric(s) ||
      !(length(s) == 1 || length(s) == length(contcols)) ||
      any(s <= 0 & s != -1)) {
    stop("'s' must be either a single numeric value (-1 for automatic selection or a positive value) or a numeric vector with positive values matching the number of 'contcols'.")
  }
  
  # Check special case of alpha = 0 (DIBmix) or alpha = 1 (IBmix)
  if (alpha == 1){
    message('alpha = 1; running IBmix.')
    best_clust <- IBmix(X, ncl, beta, catcols, contcols, randinit,
                        lambda, s, scale,
                        maxiter, nstart,
                        verbose)
  } else if (alpha == 0){
    message('alpha = 0; running DIBmix - value of beta is ignored.')
    best_clust <- DIBmix(X, ncl, catcols, contcols, randinit,
                         lambda, s, scale,
                         maxiter, nstart,
                         verbose)
  } else {
    X <- data.frame(X)
    X[, catcols] <- preprocess_cat_data(X[, catcols])
    if (scale){
      X[, contcols] <- preprocess_cont_data(X[, contcols])
    }
    
    bws_vec <- compute_s_lambda(X, contcols, catcols, s, lambda)
    
    # Construct joint density with final bandwidths
    pxy_list <- coord_to_pxy_R(X, s = bws_vec[contcols],
                               cat_cols = catcols, cont_cols = contcols,
                               lambda = bws_vec[catcols])
    
    py_x <- pxy_list$py_x
    px <- pxy_list$px
    pxy <- pxy_list$pxy
    hy <- pxy_list$hy
    
    ######################################################
    best_clust <- GIBmix_iterate(X, ncl = ncl, beta = beta, alpha = alpha,
                                 randinit = randinit,
                                 tol = 0, py_x, hy, px, maxiter,
                                 bws_vec, contcols, catcols,
                                 runs = nstart, verbose = verbose)
    ######################################################
  }
  return(best_clust)
}