R/methods.R

In dplbnDE: Discriminative Parameter Learning of Bayesian Networks by Differential Evolution

#' Print main results of evolution
#'
#' @param x A list of structures with parameters learned by \code{\link{DE}}.
#' @param \ldots Not used.
#' @return Nothing. Side-effect: print results.
#' @export
#' @keywords internal
#'
print.DE <- function(x, ...){
  cat("Number of evaluations: \t", x$N.evals, "\n")
  cat("Final population size: \t", length(x$pobFinal), "\n\n")
  cat("Summary results of fitness in final population: \n\n")
  cat("Best CLL: \t", x$BestCLL, "\n")
  cat("Worst CLL: \t", min(x$CLLPobFinal), "\n")
  cat("Median: \t", median(x$CLLPobFinal), "\n")
  cat("Std. Dev.: \t", sd(x$CLLPobFinal), "\n")
}

#' Plot main results of evolution
#' @importFrom graphics par hist
#'
#' @param x A list of structures with parameters learned by \code{\link{DE}}.
#' @param \ldots further arguments passed to plot.
#' @return Nothing. Side-effect: plots graphs.
#' @export
plot.DE <- function(x, ...){
  oldpar <- par(no.readonly = TRUE)    # Save Graphical Parameters
  par(mfrow = c(1, 2))                    # Change Graphical Parameters
  hist(x$CLLPobFinal, xlab = "CLL",  main = "CLL of final population", breaks = sqrt(length(x$pobFinal)))
  plot(x$evaluations, x$convergence, type = "l", xlab = "Evaluations", ylab = "CLL",
       main = "Convergence plot")
  on.exit(par(oldpar))            # Reset Graphical Parameters
}