plot.func.ert: Plot ERT Functions
In thomasWeise/plotteR: An R Package with Utilities for Graphics and Plots

Description Usage Arguments Value Examples

Plot the Expected Running Time (ERT) functions for a couple of different datasets.

## S3 method for class 'func.ert'
plot(x, goal.dim = 2L, time.dim = 1L,
  time.type = as.integer, goal.type = as.numeric,
  goal.min = goal.type(NA_real_), goal.max = goal.type(NA_real_),
  time.max = time.type(NA_real_), extract.run = identity,
  extract.runs = identity, comparator = `<=`,
  lineTypeFun = lineTypes.distinct, colorFun = colors.distinct,
  goal.markers = c(0.2, 0.4, 0.6, 0.8), time.markers = NULL,
  legend = NULL, legend.pos = "bottomleft", legend.cex = NA, ...)

`x`	the data, maybe a list of [lists of matrices or a list of vectors/lists], where each element has at least the `goal.dim` and `time.dim` dimension.
`goal.dim`	the dimension where where the goal values can be found
`time.dim`	the dimension where the time values can be found
`time.type`	the type function the time dimension, should be `as.integer` or `as.numeric`
`goal.type`	the type function the goal dimension, should be `as.integer` or `as.numeric`
`goal.min`	the minimum goal value or `NA` to pick the minimum time value
`goal.max`	the maximum goal value or `NA` to pick the minimum time value
`time.max`	the maximum time value to be used, or `NA` to pick the maximum time value of any run
`extract.run`	a function which can be used to extract the single runs from the run sets, e.g., `identity`
`extract.runs`	a function which can be used to extract the run sets from `x`, e.g., `identity`
`comparator`	the comparator, usually `<=` or `>=`
`lineTypeFun`	the line type function, a function returning the line types to use for a provided number of inputs
`colorFun`	the colors function, a function returning the color list to use for a provided number of inputs
`goal.markers`	markers for the goal values
`time.markers`	markers for the time values
`legend`	the legend names
`legend.pos`	the legend position (optional, default: bottomleft)
`legend.cex`	the character sizing for the legend, optional, default `NA`
`...`	Arguments passed on to `graphics::plot` x the coordinates of points in the plot. Alternatively, a single plotting structure, function or any R object with a `plot` method can be provided. y the y coordinates of points in the plot, optional if `x` is an appropriate structure.

a list(x=c(goal.min, goal.max), y=c(time.min, time.max)) of the minimum and maximum time and goal values encountered

library("plotteR");

set.seed(10000L);

# create a single run
make.run <- function(q.best, time.worst, ...) {
  repeat {
    x <- sort(unique(as.integer(c(1L, time.worst, runif(n=19L, min=1L, max=time.worst)))));
    if(length(x) == 21L) { break; }
  }

  range <- 1 - q.best;
  repeat {
    l <- unique((exp(runif(n=20L, min=1L, max=5L))-exp(1L))/exp(5L));
    if(length(l) != 20L) { next; }
    if(any((l<=0) | (l >= 1))) { next; }
    y <- sort(unique(c(q.best, q.best + (range*l))), decreasing = TRUE);
    if(length(y) == 21L) { break; }
  }
  return(matrix(c(x, y), ncol=2L))
}

# make n runs
make.runs <- function(n, q.best, time.worst) {
  return(lapply(X=seq_len(n), FUN=make.run, q.best=q.best, time.worst=time.worst));
}

# plot five example ERTs
plot.func.ert(x = list(make.runs(20, 0.0, 1e6),
                       make.runs(20, 0.5, 1e6),
                       make.runs(20, 0.0, 1e4),
                       make.runs(20, 0.5, 1e4),
                       make.runs(20, 0.1, 1e5)),
              legend=c("slow+good", "slow+bad", "fast+good", "fast+bad", "ok+ok"),
              log="y",
              time.max=1e6,
              goal.min=0,
              goal.max=1,
              goal.markers = c(0.1, 0.5),
              time.markers = c(1e6, 1e4, 1e5));