DispersionRatio: Dispersion Ratio

In xegaSelectGene: Selection of Genes and Gene Representation Independent Functions

Dispersion Ratio

Description

The dispersion ratio is computed as the ratio DM(t)/DM(k) where DM(t) is the dispersion measure of period t and DM(k) the dispersion measure of period max(1, (t-k)). k is specified by lF$ScalingDelay().

Usage

DispersionRatio(popStat, DM, lF)

Arguments

popStat	Population statistics.
DM	Dispersion function.
lF	Local configuration.

Details

The dispersion ratio may take unreasonably high and low values leading to numerical underflow or overflow of fitness values. Therefore, we use hard thresholding to force the dispersion ratio into the interval [lF$DRmin(), lF$DRmax()]. The default interval is [0.5, 2.0].

Value

Dispersion ratio.

See Also

Other Scaling: ContinuousScaleFitness(), ScaleFitness(), ScalingFitness(), ThresholdScaleFitness()

Examples

p<-matrix(0, nrow=3, ncol=8)
p[1,]<-c(14.1,  0.283,  5.53, 14.0, 19.4, 38.1, 90.2, 6.54)
p[2,]<-c(20.7,  0.794, 14.63, 19.0, 26.5, 38.8, 71.4, 5.27)
p[3,]<-c(24.0,  6.007, 16.89, 24.1, 29.2, 38.8, 73.4, 6.50)
F<-list()
F$ScalingDelay<-function() {1}
F$DRmax<-function() {2.0}
F$DRmin<-function() {0.5}
dm<-DispersionMeasureFactory("var")
DispersionRatio(p, dm, F)
F$ScalingDelay<-function() {2}
DispersionRatio(p, dm, F)

xegaSelectGene documentation built on April 16, 2025, 5:12 p.m.