tester.batch: Tester Batch

In caranha/EvoGCP: Evolutionary Computation Tools for the Graph Coloring Problem

Description

Solves a large number of 3-gcp problems, with a specified set of solver and solver parameters, and a specified set of graph parameters. Results are saved as the tester goes along, and it is able to load previous results to avoid re-computation.

Usage

tester.batch(solverpar.list, graphpar.list, results.file, graphs.file, nfe,
  random.seed = 42, ncores = Inf, output = "none")

Arguments

solverpar.list	a list where each element is a parameter list for test.solver
graphpar.list	a list where each element is a parameter list for dataset
results.file	a filename for saving partial and final results
graphs.file	a filename for saving partial and final data sets
nfe	number of maximum function evaluations for this experiment
random.seed	the random number seed for this experiment
ncores	the maximum number of cores to be used in this experiment

Value

Nothing, but saves the files "results.file" and "graphs.file" with the calcultion results

Examples

graphpar.list = list()
graphpar.list[["random90"]] = list(name = "random90", size = 5, nodes = 90, density = 3, method = "random")
graphpar.list[["minton90"]] = list(name = "random90", size = 5, nodes = 90, density = 3, method = "minton")

solverpar.list = list()
solverpar.list[["cs.toda"]] = list(name = "cs.toda", method = "cuckoo", pop = 10, pc = 0.0001, compare = F, policy = "levy", alpha = 1, beta = 1.5, E = NULL)
solverpar.list[["abc.togashi"]] = list(name = "abc.togashi", method = "abc", pop = 500, onlooker = 500, c = 3, scout = 1, limit = 300)

tester.batch(solverpar.list, graphpar.list, results.file = "foo.Rda", graphs.file = "bar.Rda", nfe = 10000, random.seed = 42, ncores = 1)

file.remove("foo.test")
file.remove("bar.test")

caranha/EvoGCP documentation built on May 3, 2021, 3:40 p.m.