SPlus Cluster Analysis

[Nina Bhanap <NBhanap@Zefer.com>]

Hi there

I am brand new to S-Plus and have four questions regarding Statistics -> Cluster Analysis -> Agglomerative Hierarchical:

1) The results I get in the Report window... how do I interprete them?
o Call
o Merge
o Order of objects
o Height
o AC
o Available arguments

2) How do I figure out HOW these objects have been clustered... variable
characteristics, means, etc.

3) What if I have some discrete, some continuous and some binary variables;
for example:

o Credit balance:                       ($0.00 to $1,000,000.00)
o Transaction Preference:       Online/Offline
o Revenue bracket:              1, 2 ,3 or 4

How would I sensibly use these variables?

4) Are there tools/techniques in splitting the tree horizontally so that I
can say:

        a) I want 3-5 clusters
  OR
        b) split the tree visually and then let the program tell me what my
clusters are, given that split

I would very much appreciate any help you can give.

Thanks in advance - Nina

:::.:::..::::.:.:::.:::.::::

//nina.bhanap
//business.strategy
//zefer.san francisco
//vox.415.762.3487
//fax.415.247.9929

:::::.:::::::.:::.:::.:::::