| EvalGeneU {xegaSelectGene} | R Documentation |
Evaluates a gene in a problem environment
Description
EvalGeneU evaluates a gene in
a problem environment.
Usage
EvalGeneU(gene, lF)
Arguments
gene |
A gene. |
lF |
The local configuration of the genetic algorithm. |
Details
If the evaluation of the fitness function of the problem environment fails, the following strategy is used: We catch the error and print it, ignore it:
The error handler returns NA.
We check for the error and update the gene:
-
$evaluatedTRUE. -
$evalFailTRUE. -
$fitis set to the minimum fitness in the population.
The boolean function lF$ReportEvalErrors controls
the output of error messages for evaluation failures.
Rationale: In grammatical evolution, the standard approach
ignores attempts the evaluate incomplete programs.
Value
A gene (with $evaluated==TRUE).
Future improvement
Provide configurable error handlers. Rationale: Make debugging for new problem environments easier. Catch communication problems in distributed/parallel environments.
See Also
Other Evaluation Functions:
EvalGeneDet(),
EvalGeneR(),
EvalGeneStoch(),
EvalGene()
Examples
Parabola2D<-Parabola2DFactory()
lF<-NewlFevalGenes(Parabola2D)
g1<-list(evaluated=FALSE, fit=0, gene1=c(1.0, -1.5, 3.37))
g2<-list(evaluated=FALSE, fit=0, gene1=c(0.0, 0.0, 0.0))
EvalGeneU(g1, lF)
EvalGeneU(g2, lF)