nonDominated: nonDominated

In tudob/wfg: WFG toolkit implementation package

Description

This determines the non-dominated points.
nonDominated is similar to emoa::nonDominated_points() but it retains extra columns.
This allows to later know the search-space coordinates of the returned points.

nonDominate is also similar to mco::paretoSet() however that depends on the optimization-algorithm being able to calculate the logical dominated-column itself. nonDominated does not have any such dependency.

Usage

nonDominated(matr, numCoords)

Arguments

matr	Matrix with an individual in each row. The meanings of columns are determined by numCoords:
numCoords	The first numCoords columns are usually the search-space coordinates of the individuals (other aspects can be stored as well), the remaining columns are the objective-space values of the individuals with which domination will be determined.

Value

the matrix of non-dominated individuals

Examples

# this shows two values on the two axis and their two search-space coordinates as color and form
value1 = runif(20)
value2 = runif(20)
coord1 = 1:20
coord2 = 20:1
d = cbind(coord1, coord2, value1, value2)
d
nonDom = nonDominated(d, 2)
nonDom
plot(d[,3:4], col=d[, 1], pch=d[, 2], xlim=c(0,1), ylim=c(0,1))
plot(nonDom[,3:4], col=nonDom[, 1], pch=nonDom[, 2], xlim=c(0,1), ylim=c(0,1))

tudob/wfg documentation built on June 1, 2019, 2:54 a.m.