xegaGpFilterCrossGene {xegaGpGene}

Crossover of 2 derivation tree genes with node filter.

Description

xegaGpFilterCrossGene() swaps two randomly extracted subtrees between 2 genes. Subtrees must have the same root in order to be compatible. The current implementation performs at most lF$maxtrials() attempts to find compatible subtrees. If this fails, the original gene is returned. Only nodes with a depth between lF$MinMutInsertionDepth() and lF$MaxMutInsertionDepth() are considered as candidate roots of derivation trees to be swapped by crossover.

Usage

xegaGpFilterCrossGene(ng1, ng2, lF)

Arguments

ng1

Derivation tree.

ng2

Derivation tree.

lF

Local configuration of the genetic algorithm.

Details

Crossover is controlled by three local parameters:

• lF$MinCrossDepth() and lF$MaxCrossDepth() control the possible exchange points for subtrees. The depth of the exchange node must be between lF$MinMutInsertionDepth() and lF$MaxMutInsertionDepth().

• lF$MaxTrials(): Maximal number of trials to find compatible subtrees. If compatible subtrees are not found, the gene is returned unchanged.

Value

List of 1 derivation tree.

See Also

Other Crossover: xegaGpAllCross2Gene(), xegaGpAllCrossGene(), xegaGpFilterCross2Gene()

Examples

gene1<-xegaGpInitGene(lFxegaGpGene)
gene2<-xegaGpInitGene(lFxegaGpGene)
xegaGpDecodeGene(gene1, lFxegaGpGene)
xegaGpDecodeGene(gene2, lFxegaGpGene)
newgene<-xegaGpFilterCrossGene(gene1, gene2,  lFxegaGpGene)
xegaGpDecodeGene(newgene[[1]], lFxegaGpGene)