ChemoSpec (version 3.0-1)

sPlotSpectra: s-Plot of Spectra Data Post PCA

Description

sPlotSpectra produces a scatter plot of the correlation of the variables against their covariance for a chosen principal component. It allows visual identification of variables driving the separation and thus is a useful adjunct to traditional loading plots.

Usage

sPlotSpectra(spectra, pca, pc = 1, tol = 0.05, ...)

Arguments

spectra

An object of S3 class Spectra.

pca

The result of a pca calculation on Spectra (i.e. the output from classPCA or robPCA).

pc

An integer specifying the desired pc plot.

tol

A number describing the fraction of points to be labeled. tol = 1.0 labels all the points; tol = 0.05 labels the most extreme 5 percent.

...

Additional parameters to be passed to plotting functions.

Value

• A data frame containing the covariance and correlation of the selected pc for the Spectra object. A plot of the correlation vs. covariance is created.

References

Wiklund, Johansson, Sjostrom, Mellerowicz, Edlund, Shockcor, Gottfries, Moritz, and Trygg. "Visualization of GC/TOF-MS-Based Metablolomics Data for Identification of Biochemically Interesting Compounds Usings OPLS Class Models" Analytical Chemistry Vol.80 no.1 (2008): 115-122. https://github.com/bryanhanson/ChemoSpec

Examples

data(SrE.IR)
IR.pca <- classPCA(SrE.IR)
myt <- expression(bolditalic(Serenoa)~bolditalic(repens)~bold(IR~Spectra))
splot <- sPlotSpectra(spectra = SrE.IR, pca = IR.pca, pc = 1, tol = 0.005,
main = myt)