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

Description Usage Arguments Value Examples

Plot the Empirical Cumulative Distribution Functions (ECDFs) for a couple of different datasets.

## S3 method for class 'func.ecdf'
plot(x, goal = 0, goal.dim = 2L, time.dim = 1L,
  time.type = as.integer, time.min = time.type(1L),
  time.max = time.type(NA_real_), goal.min = 0, goal.max = 1,
  extract.runs = identity, comparator = `<=`, extract.run = identity,
  lineTypeFun = lineTypes.distinct, colorFun = colors.distinct,
  goal.markers = c(0.2, 0.4, 0.6, 0.8), time.markers = NULL,
  legend = NULL, legend.pos = "topleft", 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`	the goal value, i.e., the value which must be reached to be counted as success
`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`
`time.min`	the minimum time value to be used for the diagram, or `NA` to use the smallest time value in any run
`time.max`	the maximum time value to be used, or `NA` to pick the maximum time value of any run
`goal.min`	the minimum goal value
`goal.max`	the maximum goal value
`extract.runs`	a function which can be used to extract the run sets from `x`, e.g., `identity`
`comparator`	the comparator, usually `<=` or `>=`
`extract.run`	a function which can be used to extract the single runs from the run sets, e.g., `identity`
`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: topleft)
`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(time.min, time.max), y=c(goal.min, goal.max)) of the minimum and maximum time and goal values encountered

library("plotteR");

set.seed(10000L);
time.max.pow <- 8;

# create a single run, where the quality dimension reaches to end
make.run <- function(end) {
  repeat {
    x <- sort(unique(as.integer(runif(n=20L, min=1L,
                     max=(10^(runif(n=1L, min=2, max=time.max.pow)))))));
    if(length(x) == 20L) { break; }
  }
  repeat {
    y <- sort(unique(c(end, runif(n=19L, min=end, max=500))), decreasing=TRUE);
    if(length(y) == 20L) {
      break;
    }
  }
  return(matrix(c(x, y), ncol=2L))
}

# make n runs where m reach below 0, i.e., whose ECDF reaches m/n
make.runs <- function(n, m) {
  return(lapply(X=seq_len(n),
                FUN=function(i) {
                  if(i <= m) { end <- runif(n=1L, min=-10L, max=0L); }
                  else       { end <- runif(n=1L, min=1L, max=100L); }
                  return(make.run(end));
                }))
}

# plot five example ECDFs, where the end results reach 3/20, 10/20, 5/20, 15/20,
# and 19/20, respectively
plot.func.ecdf(x = list(make.runs(20, 3),
                        make.runs(20, 10),
                        make.runs(20, 5),
                        make.runs(20, 15),
                        make.runs(20, 19)),
               legend=c("worst", "good", "bad", "better", "best"),
               time.markers=c(1e2, 1e4, 1e6, 1e8),
               log="x",
               time.max=(10^time.max.pow));