benchmarkGeneratorNKL: NK-Landscape Benchmark Creation

In CEGO: Combinatorial Efficient Global Optimization

NK-Landscape Benchmark Creation

Description

Function that generates a NK-Landscapes.

Usage

benchmarkGeneratorNKL(N = 10, K = 1, PI = 1:K, g)

Arguments

N	length of the bit strings
K	number of neighbours contributing to fitness of one position
PI	vector, giving relative positions of each neighbour in the bit-string
g	set of fitness functions for each possible combination of string components. Will be randomly determined if not specified. Should have N rows, and 2^(K+1) columns.

Value

the function of type cost=f(bitstring). Returned fitness values will be negative, for purpose of minimization.

Examples

fun <- benchmarkGeneratorNKL(6,2)
fun(c(1,0,1,1,0,0))
fun(c(1,0,1,1,0,1))
fun(c(0,1,0,0,1,1))
fun <- benchmarkGeneratorNKL(6,3)
fun(c(1,0,1,1,0,0))
fun <- benchmarkGeneratorNKL(6,2,c(-1,1))
fun(c(1,0,1,1,0,0))
fun <- benchmarkGeneratorNKL(6,2,c(-1,1),g=matrix(runif(48),6))
fun(c(1,0,1,1,0,0))
fun(sample(c(0,1),6,TRUE))

CEGO documentation built on May 29, 2024, 3:35 a.m.