func.ertEcdf: Create an ECDF Function over ERTs

In thomasWeise/plotteR: An R Package with Utilities for Graphics and Plots

Description

Create an Empirical Cumulative Distribution Function (ECDF) function over Expected Running Times (ERTs), i.e., a list with two membery x and y that stand for the coordinates of the function. This approach is suitable if we want to plot an ECDF over multiple distinct problems. We can compute the expected running time of an algorithm on the given problem to reach the goal. Over all the ERTs of all problems, we can then draw the ECDF.

Usage

func.ertEcdf(x, goal = 0, goal.dim = 2L, time.dim = 1L,
  comparator = `<=`, time.type = as.integer, goal.type = as.numeric,
  time.min = 1L, time.max = time.type(NA_real_),
  extract.run = identity, extract.runs = identity)

Arguments

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
comparator	the comparator, usually `<=` or `>=`
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
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
extract.run	a function which can be used to extract the single runs from the extracted run sets, e.g., identity
extract.runs	a function which can be used to extract the run sets from x, e.g., identity

Value

a list(x=c(...), y=c(...)) where the two members x and y will be vectors of corresponding coordinates

thomasWeise/plotteR documentation built on May 29, 2019, 5:41 a.m.