Elitism: Elitism

In dGAselID: Genetic Algorithm with Incomplete Dominance for Feature Selection

Description

Operator for elitism.

Usage

Elitism(results, elitism, ID)

Arguments

results	Results returned by EvaluationFunction().
elitism	Elite population in percentages.
ID	Dominance. The default value is "ID1". Use "ID2" for Incomplete Dominance.

Examples

## Not run: 
 library(genefilter)
 library(ALL)
 data(ALL)
 bALL = ALL[, substr(ALL$BT,1,1) == "B"]
 smallALL = bALL[, bALL$mol.biol %in% c("BCR/ABL", "NEG")]
 smallALL$mol.biol = factor(smallALL$mol.biol)
 smallALL$BT = factor(smallALL$BT)
 f1 <- pOverA(0.25, log2(100))
 f2 <- function(x) (IQR(x) > 0.5)
 f3 <- ttest(smallALL$mol.biol, p=0.1)
 ff <- filterfun(f1, f2, f3)
 selectedsmallALL <- genefilter(exprs(smallALL), ff)
 smallALL = smallALL[selectedsmallALL, ]
 rm(f1)
 rm(f2)
 rm(f3)
 rm(ff)
 rm(bALL)
 sum(selectedsmallALL)
 set.seed(1357)

 population0<-InitialPopulation(smallALL, 14, 8, FALSE)
 individuals0<-Individuals(population0)
 results0<-EvaluationFunction(smallALL, individuals0, response="mol.biol",
             method=knn.cvI(k=3, l=2), trainTest="LOG")
 Elitism(results0, 25, ID="ID1")
 Elitism(results0, 25, ID="ID2")
 
## End(Not run)

dGAselID documentation built on May 2, 2019, 1:27 p.m.