R/plot_runtime.R

In MarianSchoen/DMC: Comparison of different Deconvolution Models in several scenarios

#' create barplot of algorithm runtimes
#'
#' @param results.df data frame containing deconvolution results as
#' returned by prepare_data()
#' @param title character, plot title
#' @param algorithm.order character vector of algorithm names
#' specifying the plotting order
#' @return runtime plot

plot_runtime <- function(
  results.df,
  title = NULL,
  algorithm.order = NULL
) {
  # parameter checks
  if (!is.data.frame(results.df)) {
      stop("Invalid input: results.df has to be data frame
            as returned by prepare_data()")
  }
  if (!all(c("algorithm", "time", "cell_type") %in% colnames(results.df))) {
      stop("required columns missing in results.df")
  }
  if (is.null(title)) {
     title <- "runtime comparison"
  }
  if (!is.null(algorithm.order)) {
    if (!is.character(algorithm.order)) {
      stop("celltype.order must be a charcter vector")
    }
    if (!all(algorithm.order %in% unique(results.df$algorithm)) ||
        length(algorithm.order) != length(unique(results.df$algorithm))) {
      stop("algorithm.order does not fit the algorithm column of results.df")
    }
  }
  # reduce to the 'overall' rows
  results.df <- results.df[which(results.df$cell_type == "overall"), ]

  # average over all repetitions
  mean.times <- tapply(results.df$time, results.df$algorithm, mean)
  sd.times <- tapply(results.df$time, results.df$algorithm, sd)

  # create data frame containing runtimes
  runtimes <- data.frame(
    algorithm = names(mean.times),
    runtime = mean.times,
    sd = sd.times
  )
  if (!is.null(algorithm.order)) {
      runtimes$algorithm <- factor(runtimes$algorithm, levels = algorithm.order)
  }

  # plot runtime as lineplot
  runtime.plot <- ggplot(runtimes) +
  geom_bar(
      aes(
        y = algorithm,
        x = log10(as.numeric(as.character(runtime)) + 1),
        fill = algorithm
      ),
      stat = "identity", position = "dodge"
  ) +
  ylab("algorithm") + xlab("time (s)") + ggtitle(title) +
  theme(
      legend.position = "none",
      title = element_text(size = 24),
      axis.title.x = element_text(size = 22),
      axis.text.x = element_text(size = 20),
      axis.title.y = element_text(size = 22),
      axis.text.y = element_text(size = 20)
  ) +
  scale_x_continuous(
      labels = function(x) {
        round(10^as.numeric(x) - 1, 2)
      }
  )
  return(runtime.plot)
}