ScalingFactory: Scaling Factory

In xegaSelectGene: Selection of Genes and Gene Representation Independent Functions

Scaling Factory

Description

Scaling Factory

Usage

ScalingFactory(method = "NoScaling")

Arguments

method

A scaling method. Available methods are: "NoScaling": Identity (Default). "ConstantScaling": fit^k with constant exponent. Function ConstantScaling(). "ThresholdScaling": If the dispersion ratio is larger than 1+threshold, use a constant scaling exponent with a value below 1 (decrease of selection pressure). Function ThresholdScaling(). If the dispersion ratio is lower than 1-threshold, use a constant scaling exponent with a value above 1 (increase of selection pressure). Else use a scaling exponent of 1. This means no scaling. "ContinuousScaling": Use weighted dispersion ratio as scaling exponent. Function ContinuousScaling().

Value

A scaling function.

See Also

Other Configuration: DispersionMeasureFactory(), EvalGeneFactory(), SelectGeneFactory()

Examples

fit<-sample(10, 20, replace=TRUE)
lF<-list()
lF$ScalingExp<-parm(2)
Scale<-ScalingFactory()
fit
Scale(fit, lF)
Scale<-ScalingFactory("ConstantScaling")
Scale(fit, lF)

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