PLSDA.CV: PLSDA cross validation

In flajole/MApckg: R package of MetaboAnalyst functions

Description

Perform PLSDA classification and feature selection. Writes two output files: "plsda_vip.csv" and "plsda_coef.csv".

Usage

PLSDA.CV(dataSet, analSet, methodName = "CV",
  compNum = GetDefaultPLSCVComp(dataSet), choice = "Q2")

Arguments

dataSet	List, data set object generated by MS_to_MA function.
analSet	List, containing the results of statistical analysis (can be just an empty list).
methodName	Cross validation method "LOOCV" or "CV"
compNum	The maximum number of components to search.
choice	The determining performance measure, one of "Q2", "R2", "Accuracy"

Value

Native analSet with one added $plsda element consisting of:

• $best.num - the detected best number of components

• $choice - value of choice argument

• $coef.mat - matrix of VIPs (variable importance in projection)

• $vip.mat - matrix of weighted sums of absolute regression coefficients

• $fit.info - all performance measures calculated for different number of components

See Also

PLS.Anal, PLS.Loadings, PLSDA.Permut, PlotPLS

flajole/MApckg documentation built on May 16, 2019, 1:16 p.m.