This collection of gene representation-independent
mechanisms for evolutionary and genetic algorithms
for the R-package xega < https://CRAN.R-project.org/package=xega>
contains four groups of functions:
First, functions for selecting a gene
in a population of genes according to its fitness value
and for adaptive scaling of the fitness values as well as for
performance optimization and measurement offer several
variants for implementing the survival of the fittest.
Second, evaluation functions for
deterministic functions avoid recomputation.
Evaluation of stochastic functions incrementally improve
the estimation of the mean and variance of fitness values
at almost no additional cost. Evaluation functions
for gene repair handle error-correcting decoders.
Third, timing and counting functions for profiling the
algorithm pipeline are provided to assess bottlenecks in
the algorithms.
Fourth, a small collection of problem environments
for function optimization, combinatorial optimization, and
grammar-based genetic programming and grammatical evolution
is provided for tutorial examples.
For xega's architecture,
see Geyer-Schulz, A. (2025)
# The R script xegaSelectGene.R
The directory selectGene contains the implementation
of a set of selection and evaluation functions for
a genetic algorithm.
The easiest installation is by using install.packages from R.
1. Linux:
Start R in the directory where you have stored xegaSelectGene_0.0.0.1.tar.gz
> install.packages("xegaSelectGene_0.0.0.1.tar.gz")
1. Windows:
Start R in the directory where you have stored xegaSelectGene_0.0.0.1.zip
> install.packages("xegaSelectGene_0.0.0.1.zip")
The R script xegaSelectGene.R generates
1. the man pages of the package.
2. the pdf documentation of the package.
3. the compilation and installation of the package.
(You must adapt the variables path and pName
in the script selectGene.R:
path<-"~/dev/iaa/genetic/xega/"
pName<-"xegaSelectGene"
to the directory where you copy the package.
You need to install the R-package devtools for this R script.
There are two ways to use the code:
1. Change into the R directory which contains the code, start R and load all code files:
source("xegaSelectGene.R")
Note that the package selectGene provides selection and evaluation functions
to all genetic algorithms of the family sgX:
sga, sgp, sgPerm, (and sge, sgde).
Use the pdf-Manual!
2. Adapt the path in selectGene.R and install the library selectGene:
Rscript xegaSelectGene.R
Restart R and
library(xegaSelectGene)
You get all elements of the package by
xegaSelectGene::
and pressing the tabulator.
You get the man page by e.g
SelectLinearRank?
or
help(SelectLinearRank)
All examples given on a man page can be executed by
example(SelectLinearRank)
Note that man pages and examples are only available, if you install
the package as a library.