Skip to contents

Create a new Visualizer.


  .name = NULL,
  .rmd_options = list(),
  .rmd_show = TRUE,



The visualization function.


(Optional) The name of the Visualizer.


(Optional) List of options to control the aesthetics of the Visualizer's visualization in the knitted R Markdown report. Currently, possible options are "height" and "width" (in inches). The argument must be specified by position or typed out in whole; no partial matching is allowed for this argument.


If TRUE (default), show the resulting visualization in the R Markdown report; if FALSE, hide output in the R Markdown report.


Arguments to pass into .viz_fun().


A new instance of Visualizer.


When visualizing or running the Experiment (see Experiment$visualize() or Experiment$run()), the named arguments fit_results, eval_results, and vary_params are automatically passed into the Visualizer function .viz_fun() and serve as placeholders for the Experiment$fit() results, the Experiment$evaluate() results, and the name of the varying parameter, respectively. To visualize the performance of a method(s) fit and/or its evaluation metrics then, the Visualizer function .viz_fun() should take in the named arguments fit_results and/or eval_results. See Experiment$fit() or fit_experiment() for details on the format of fit_results. See Experiment$evaluate() or evaluate_experiment() for details on the format of eval_results. If the Visualizer is used for Experiments with varying parameters, vary_params should be used as a stand in for the name of this varying parameter.


# create an example Visualizer function
power_plot_fun <- function(fit_results, vary_params = NULL, col = "X1") {
  if (!is.null(vary_params)) {
    # deal with the case when we vary across a parameter that is vector-valued
    if (is.list(fit_results[[vary_params]])) {
      fit_results[[vary_params]] <- list_col_to_chr(fit_results[[vary_params]],
                                                    name = vary_params,
                                                    verbatim = TRUE)
  plt <- ggplot2::ggplot(fit_results) +
    ggplot2::aes(x = .data[[paste(col, "p-value")]],
                 color = as.factor(.method_name)) +
    ggplot2::geom_abline(slope = 1, intercept = 0, 
                         color = "darkgray", linetype = "solid", size = 1) +
    ggplot2::stat_ecdf(size = 1) +
    ggplot2::scale_x_continuous(limits = c(0, 1)) +
    ggplot2::labs(x = "t", y = "P( p-value \u2264 t )", 
                  linetype = "", color = "Method")
  if (!is.null(vary_params)) {
    plt <- plt + ggplot2::facet_wrap(~ .data[[vary_params]])

# create Visualizer using the default arguments (i.e., col = "X1")
power_plot1 <- create_visualizer(.viz_fun = power_plot_fun, 
                                 .name = "Power Plot (X1)")
# create Visualizer using non-default arguments (i.e., col = "X2")
power_plot2 <- create_visualizer(.viz_fun = power_plot_fun,
                                 .name = "Power Plot (X2)",
                                 # additional named parameters to pass to power_plot_fun()
                                 col = "X2")

# create Visualizer from a function in the built-in Visualizer library
pred_err_plot <- create_visualizer(.viz_fun = plot_pred_err,
                                   .name = "Prediction Error Plot",
                                   # additional named parameters to pass to plot_pred_err()
                                   truth_col = "y", estimate_col = "predictions")
# change figure height/width when displaying Visualizer in Rmd report
pred_err_plot <- create_visualizer(.viz_fun = plot_pred_err,
                                   .name = "Prediction Error Plot",
                                   .rmd_options = list(height = 8, width = 12),
                                   # additional named parameters to pass to plot_pred_err()
                                   truth_col = "y", estimate_col = "predictions")