evaluateMissingCandidateSolutions: evaluateMissingCandidateSolutions

In spotGUI: Graphical User Interface for the Package 'SPOT'

Description

evaluateMissingCandidateSolutions evaluates all non-evaluated candidate solutions in a gived data.frame. This function is used as a convenience function for codes that are automatically generated by the SPOT-GUI.

Usage

evaluateMissingCandidateSolutions(currX, currY = NULL, fun)

Arguments

currX	A matrix containing all candidate solutions. One candidate per row.
currY	A column vector with all known objective function results for the given matrix of candidate solutions. Default = NULL (In this case all candidate solutions will be evaluated). Missing values have to be marked as NA.
fun	The objective function on which the given candidate solutions shall be evaluated.

Value

y An updated column vector with evaluation results for all candidate soltuions given in currX

Examples

library(SPOT)
spotData <- NULL
#Generating DOE
spotData$x <- designLHD(x = NULL, lower = c(-5, -5), upper = c(5, 5),
                        control = list(size = 10,
                             types = c("numeric", "numeric")))

#Evaluating Candidate Solutions
spotData$y <- evaluateMissingCandidateSolutions(
        currX = spotData$x, currY = spotData$y, fun = funSphere)

#Build model on evaluated data
spotData$modelFit <- buildKriging(as.matrix(spotData$x),as.matrix(spotData$y))

spotGUI documentation built on March 31, 2021, 1:06 a.m.