| SelectLinearRankTSR {xegaSelectGene} | R Documentation |
Linear rank selection.
Description
SelectLinearRankTSR implements selection
with interpolated target sampling rates.
Usage
SelectLinearRankTSR(fit, lF, size = 1)
Arguments
fit |
Fitness vector. |
lF |
Local configuration. |
size |
Size of return vector (default: 1). |
Details
The target sampling rate is a linear interpolation
between lF$MaxTSR and Min<-2-lF$MaxTSR,
because the sum of the target sampling rates is $n$.
The target sampling rates are computed and used as a fitness
vector for stochastic universal sampling algorithm
implemented by SelectSUS.
lF$MaxTSR should be in [1.0, 2.0].
TODO: More efficient implementation. We use two sorts!
Value
The index vector of selected genes.
References
Grefenstette, John J. and Baker, James E. (1989): How Genetic Algorithms Work: A Critical Look at Implicit Parallelism In Schaffer, J. David (Ed.) Proceedings of the Third International Conference on Genetic Algorithms on Genetic Algorithms, pp. 20-27. (ISBN:1-55860-066-3)
See Also
Other Selection Functions:
SelectDuel(),
SelectLRSelective(),
SelectPropFitDiffM(),
SelectPropFitDiffOnln(),
SelectPropFitDiff(),
SelectPropFitM(),
SelectPropFitOnln(),
SelectPropFit(),
SelectSTournament(),
SelectSUS(),
SelectTournament(),
SelectUniformP(),
SelectUniform()
Examples
fit<-sample(10, 15, replace=TRUE)
SelectLinearRankTSR(fit, NewlFselectGenes())
SelectLinearRankTSR(fit, NewlFselectGenes(), length(fit))