Algorithm Interface to ES + GLG

Share:

Description

Interface for mixed model tuning of an Evolution Strategy ES optimizing Gaussian Landscapes. For each repeat of an ES parameter setting, a different seed for the ES is used. At the same time, this seed is used to create different Gaussian Landscapes with the Gaussian Landscape Generator GLG, see spotGlgCreate.

Usage

1

Arguments

spotConfig

Contains the list of spot configurations, results of the algorithm can be passed to this list instead of the .res file. spotConfig defaults to "NA", and will only be passed to the Algorithm if spotConfig$spot.fileMode=FALSE. See also: spotGetOptions Items used are:

alg.currentDesign: data frame holding the design points that will be evaluated
io.apdFileName: name of the apd file
io.desFileName: name of the des file
io.resFileName: name of the res file, for logging results (if spotConfig$spot.fileMode==TRUE)
spot.fileMode: boolean, if selected with true the results will also be written to the res file, otherwise it will only be saved in the spotConfig returned by this function

Value

this function returns the spotConfig list with the results in spotConfig$alg.currentResult

See Also

SPOT spot demo optim spotFuncStartBranin spotAlgStartSannVar

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.