clustering algorithm detail

Hi all,

I wanted to run the hclust (or any other clustering algorithm) on a distance matrix. I have formed the distance matrix as:

distmat:

        a        b         c         d        e
a    0.00    0.96    1.60    1.60    1.68
b    0.96    0.00    0.96    1.80    2.64
c   1.60    0.96    0.00    0.84    1.80
d   1.60    1.80    0.84    0.00    0.96
e   1.68    2.64    1.80    0.96    0.00

Now, I would like to run a clustering algorithm on it. I tried:

newclust = hclust(distmat)

and got the following error:

Error in if (n < 2) stop("Must have n >= 2 objects to cluster") :
        argument is of length zero

I understand that the documentation says that the matrix ' d: a dissimilarity structure as produced by 'dist''. Does that mean that I need to convert it into a lower triangle matrix? I also tried:

newclust = hclust(vech(distmat))

but this gave a similar error. Where am I going wrong?


Also, will hclust be able to handle 'NA' in the dissimilarity matrix?

many thanks,


Tim




---------------------------------

	[[alternative HTML version deleted]]