plsda: Partial Least Squares Discriminant Analysis

Description

plsda is used to fit PLS models for classification.

Usage

plsda(x, ...)
## S3 method for class 'default':
plsda(x,  y, ncomp = 2, ...)
predict.plsda(object, newdata = NULL, ncomp = NULL, type = "class", ...)

Arguments

a matrix or data frame of predictors

a factor or indicator matrix for the discrete outcome. If a matrix, the entries must be either 0 or 1 and rows must add to one

ncomp

the number of components to include in the model

...

arguments to pass to plsr (code{plsda} only)

object

an object produced by plsda

newdata

a matrix or data frame of predictors

type

either "class", "prob" or "raw" to produce the predicted class, class probabilities or the raw model scores, respectively.

Value

For plsda, an object of class "plsda" and "mvr". The predict method produces either a vector, matrix or three-dimensional array, depending on the values of type of ncomp. For example, specifying more than one value of ncomp with type = "class" with produce a three dimensional array but the default specification would produce a factor vector.

Details

If a factor is supplied, the appropriate indicator matrix is created by plsda.

A multivariate PLS model is fit to the indicator matrix using the plsr function.

To predict, the softmax function is used to normalize the model output into probability-like scores. The class with the largest score is the assigned output class.

Examples

Run this code

data(mdrr)

tmpX <- scale(mdrrDescr)

plsFit <- plsda(tmpX, mdrrClass, ncomp = 3)

table(predict(plsFit), mdrrClass)