hca.nlda: Hierarchical Clustering for Class 'nlda'

Description Usage Arguments Author(s) See Also Examples

Description

Hierarchical clustering based on the Mahalanobis distances between group centers for class nlda. Group centers coordinates of the training data points can be calculated using all possible discriminant functions or a selected number of discriminant functions.

Usage

1
2
3
4
  hca.nlda(x, method="complete",df2use=NULL,
           main="Aggregation of group centres",
           ylab="Mahalanobis distance",
           xlab="", sub="",...)

Arguments

x

An object of class nlda.

method

Agglomeration method to be used. This should be an unambiguous abbreviation of "ward", "single", "complete", "average", "mcquitty", "median" or "centroid".

df2use

Discriminant functions to be included in the HCA (by default all of DFs are considered if df2use=NULL).

main, sub, xlab, ylab

Character strings for annotating the HCA plot. For details, see plclust.

...

Additional arguments to plclust. For details, see plclust.

Author(s)

David Enot [email protected] and Wanchang Lin [email protected].

See Also

nlda, predict.nlda

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
## load abr1
data(abr1)
cl   <- factor(abr1$fact$class)
dat <- preproc(abr1$pos , y=cl, method=c("log10","TICnorm"),add=1)[,110:500]  

## build nlda model
model    <- nlda(dat,cl)

## HCA using all DFs
hca.nlda(model)

## or only using the first 2 DFs
hca.nlda(model,df2use=1:2)

wilsontom/FIEmspro documentation built on Feb. 19, 2018, 9:03 a.m.