plot.nlda: Plot Method for Class 'nlda'

In wilsontom/FIEmspro: Flow In-jection Electrospray Mass Spectrometry Processing: \\ data processing, classification modelling and variable selection in metabolite fingerprinting

Description

Plots a set of data on one, two or more linear discriminants.

Usage

## S3 method for class 'nlda'
plot(x, panel = panel.nlda, cex = 0.7, dimen,
     abbrev = FALSE, ...)

Arguments

x	An object of class nlda.
panel	The panel function used to plot the data.
cex	Graphics parameter cex for labels on plots.
dimen	The index of linear discriminants to be used for the plot.
abbrev	Whether the group labels are abbreviated on the plots. If abbrev > 0 this gives minlength in the call to abbreviate.
...	Additional arguments to plot.

Details

This function is a method for the generic function plot() for class nlda. The behaviour is determined by the value of dimen. For the length of dimen is greater than 2, a pairs plot is used. For the length of dimen is equal to 2, a scatter plot is drawn. Otherwise, a set of histograms or density plots are drawn.

Author(s)

Wanchang Lin [e-mail] and David Enot [e-mail].

See Also

nlda, predict.nlda, hca.nlda

Examples

## load abr1
data(abr1)
cl   <- factor(abr1$fact$class)
dat <- preproc(abr1$pos , y=cl, method=c("log10","TICnorm"),add=1)[,110:500]  

## build model on all the data available
model    <- nlda(dat,cl)

## Plot second component versus first
plot(model,dimen=c(1,2),main = "Training data",abbrev = TRUE)

## Pairwise scatterplots of several components 
plot(model,main = "Training data",abbrev = TRUE)

wilsontom/FIEmspro documentation built on May 4, 2019, 6:28 a.m.