R help - Jun 2010 - Mahalanobis distance

I am a new R user.  i have a question about Mahalanobis distance.actually i have
300 rows and 7 columns. columns are different measurements, 300 rows are genes.
since genes can
classify into 4 categories. i used dist() with euclidean distance and cmdscale
to do MDS plot. but find out Mahalanobis distance may be
better. how do i use Mahalanobis() to generate similar dist object which i can
use MDS plot? second question is if should i calculate mean for
every categories for every measurement first and do 4*4 distance matrix, or i
should calculate pairwise distance first and then find category
means since i only care about relative position of 4 categories in MDS 
plot. Thank you very much.


      
	[[alternative HTML version deleted]]

Say X is your data matrix with the variable, then you could do :
 X <- matrix(rnorm(2100),300,7)
 S <- var(X)
 dist <- as.dist(
     apply(X,1,function(i){
         mahalanobis(X,i,S)
         }
     )
 )

Cheers
Joris

On Tue, Jun 22, 2010 at 11:41 PM, yoo hoo <freesuccess2001 at yahoo.com>
wrote:
> I am a new R user. ?i have a question about Mahalanobis distance.actually i
have 300 rows and 7 columns. columns are different measurements, 300 rows are
genes. since genes can
> classify into 4 categories. i used dist() with euclidean distance and
cmdscale to do MDS plot. but find out Mahalanobis distance may be
> better. how do i use Mahalanobis() to generate similar dist object which i
can use MDS plot? second question is if should i calculate mean for
> every categories for every measurement first and do 4*4 distance matrix, or
i should calculate pairwise distance first and then find category
> means since i only care about relative position of 4 categories in MDS
> plot. Thank you very much.
>
>
>
> ? ? ? ?[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>


-- 
Joris Meys
Statistical consultant

Ghent University
Faculty of Bioscience Engineering
Department of Applied mathematics, biometrics and process control

tel : +32 9 264 59 87
Joris.Meys at Ugent.be
-------------------------------
Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php