R/OptPath_plotEAF.R

In ParamHelpers: Helpers for Parameters in Black-Box Optimization, Tuning and Machine Learning

#' @title Plots attainment functions for data stored in multiple OptPaths.
#'
#' @description
#' Can be used to plot OptPaths where information for bi-objective
#' evaluation was logged for repeated runs of different algorithmic runs.
#' Pretty directly calls [eaf::eafplot()].
#'
#' @param opt.paths (`list`)\cr
#'   List of list of `OptPath` objects. First index is the algorithm / major
#'   variation in the experiment, second index is the index of the replicated
#'   run.
#' @param xlim (`numeric(2)`)\cr
#'   The x limits (x1, x2) of the plot.
#' @param ylim (`numeric(2)`)\cr
#'   The y limits (y1, y2) of the plot.
#' @param ... (any)\cr
#'   Passed on to [eaf::eafplot()].
#' @note
#'   We changed the defaults of [eaf::eafplot()] in the following way: The axis
#'   are labeled by `y.names`, colors are set to our favorite grey values and
#'   linetypes changed, too. With our colors / linetypes default it is possible
#'   to distinguish 6 different algorithms. But this can again be overwritten by
#'   the user.
#' @return [`data.frame`]
#'   Invisibly returns the data passed to [eaf::eafplot()].
#' @export
plotEAF = function(opt.paths, xlim = NULL, ylim = NULL, ...) {

  requirePackages("eaf", why = "plotEAF")

  # we need a list of lists with optimization pathes
  assertList(opt.paths, min.len = 1L, types = "list", names = "unique")

  if (!is.null(xlim)) {
    assertNumeric(xlim, len = 2L)
  }
  if (!is.null(ylim)) {
    assertNumeric(ylim, len = 2L)
  }

  algos = names(opt.paths)
  y.names = NULL
  minimize = NULL
  data = data.frame()
  for (i in seq_along(algos)) {
    # extract opt path list for current algorithm
    a = algos[i]
    runs = opt.paths[[i]]
    assertList(runs, types = "OptPath", min.len = 1L)

    # combine all fronts for this algo + add algo / repl + do some sanity checks
    fronts = lapply(seq_along(runs), function(j) {
      run = runs[[j]]
      df = as.data.frame(getOptPathParetoFront(run), stringsAsFactors = TRUE)
      cns = colnames(df)
      if (length(cns) != 2L) {
        stopf("Must always have 2 objectives in opt path. But found: %i", length(cns))
      }
      if (i == 1L && j == 1L) {
        y.names <<- cns
        minimize <<- run$minimize
      }
      if (!all(y.names == cns)) {
        stopf("Must always have the same 2 objectives in opt path: %s (first ones taken). But found here: %s",
          collapse(y.names), collapse(cns))
      }
      if (!all(minimize == run$minimize)) {
        stopf("Must always have the same 'minimize' settings for objectives in opt path: %s (first one taken).
          But found here: %s", collapse(minimize), collapse(run$minimize))
      }
      # add column for algorithm and replication
      cbind(df, .algo = a, .repl = j, stringsAsFactors = TRUE)
    })
    fronts = do.call(rbind, fronts)
    data = rbind(data, fronts)
  }
  yn1 = y.names[1L]
  yn2 = y.names[2L]
  # build parameter list for eafplot function
  f = as.formula(sprintf("%s + %s ~ .repl", yn1, yn2))
  defaults = list(
    xlab = yn1, ylab = yn2,
    percentiles = 50,
    col = c("darkgrey", "darkgrey", "darkgrey", "black", "black", "black"),
    lty = c("solid", "dashed", "dotdash", "solid", "dashed", "dotdash")
  )
  args = list(...)
  args = insert(defaults, args)
  args$data = data
  args$groups = quote(.algo)
  args$maximise = !minimize
  args$xlim = xlim
  args$ylim = ylim
  do.call(eaf::eafplot, c(list(f), args)) # due to not so good programming in eafplot the first argument can not be named if it's a formula
  return(data)
}