[R] non-parametric discriminant analysis

[Walter Durka]

Hi,

I used linear discriminant analysis (lda) to classify and to 
cross-validate samples of two plant species based on morphometric data 
and to identify the variables that best discriminate between the two 
species.
Because some of the variables are not, and can not be transformed to be, 
normally distributed, I would like to use a non-parametric method.

Are there any R packages that provide methods for non-parametric 
discriminant analysis? Would randomForest 
(http://cran.r-project.org/web/packages/randomForest) be appropriate and 
recommended?

Thanks for help and best regards

Walter Durka


-- 

*****
Dr. Walter Durka
Department Biozönoseforschung
Department of community ecology

Helmholtz-Zentrum für Umweltforschung GmbH - UFZ
Helmholtz Centre for Environmental Research - UFZ
Theodor-Lieser-Str. 4 / 06120 Halle / Germany

walter.durka at ufz.de / http://www.ufz.de/index.php?en=798
phone +49 345 558 5314 / Fax +49 345 558 5329