R/checkSample.R

In hdMTD: Inference for High-Dimensional Mixture Transition Distribution Models

#' Checks a sample
#'
#' Checks if a sample is a suitable argument for some functions within the package.
#'
#' @param X  A vector, a single-column data frame, a list, or a matrix with a
#' single row or a single column. Must be composed by nonnegative integers.
#'
#' @return Returns the sample as a vector or identifies any possible sample problems.
#'
checkSample <- function(X){
  # Allow matrices with a single column or a single row
  if (is.matrix(X)) {
      if (ncol(X) == 1 || nrow(X) == 1) {
          X <- as.vector(X)  # Convert to vector
      } else {
          stop("X must be a vector, a single-column data frame, a list, or a matrix
               with a single row or a single column.")
      }
  }

  # Convert single-column data frame to vector
  if (is.data.frame(X)) {
      if (ncol(X) != 1) stop("X must be a single chain; multiple columns are not accepted.")
      if (nrow(X) <= 1) stop("Insufficient sample size.")
      X <- as.vector(X[, 1])
  }

  # Convert list to vector if necessary
  if (is.list(X)) {
      X <- unlist(X)
  }

  # Ensure X is a simple numeric vector
  X <- as.vector(X)

  # Basic validation checks
  if (length(X) <= 1) stop("Insufficient sample size.")
  if (!is.numeric(X)) stop("X must be a numeric dataset.")
  if (any(is.na(X))) stop("NA values are not allowed in the sample.")
  if (length(unique(X)) == 1) stop("The sample must contain at least two distinct values.")
  if (any( X%%1 != 0 ) || any(X < 0)) stop("X must be composed of nonnegative integers.")
  X
}