ZipfNorm: Zipf Normalization

In bda: Binned Data Analysis

Zipf Normalization

Description

Zipf plot based normalization.

Usage

  Zipf.Normalize(x, y, cutoff=6,optim=FALSE, method)

Arguments

x, y	data: two vectors.
cutoff	a large enought value such that the values larger than the cutoff (approximately) follows a power law distribution.
optim	Find the optimal normalization parameters if TRUE
method	use both power transformation and scalingby default. If 'scaling' is specified, skip power transformation.

Value

x	reference profile (not normalized)
y	normalized profile
scaler	Linear rescaling normalization parameter estimate
power	power transformation parameter estimate
scaler.optim	Optimized estimate of the linear rescaling parameter
power.optim	Optimzed estimate of the power transformation parameter.
mat.optim	A matrix of the objective function values generated to find the optimal estimates.
coef	Coefficient table to display the estimates.

References

Wang, B. (2020) A Zipf-plot based normalization method for high-throughput RNA-Seq data. PLoS ONE, (in press).

Examples

data(LCL)
names(LCL)
x <- LCL$p47
y <- LCL$p107
outx <- ZipfPlot(x)
plot(outx,type='l')
outy <- ZipfPlot(y)
lines(outy,col=2)

out2 <- Zipf.Normalize(x,y)
outy2 <- ZipfPlot(out2$y)
lines(outy2,col=4)

  

bda documentation built on Sept. 11, 2024, 9:01 p.m.