Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: use of a distance matrix for cluster analysis in STATA

From   James Muller <[email protected]>
To   [email protected]
Subject   Re: st: use of a distance matrix for cluster analysis in STATA
Date   Mon, 20 Jun 2005 19:24:00 +1000

(please nobody flame me here :)

Just out of curiosity, why don't you use something like R for the whole process - R's language treats the sort of data objects you'll be using more generally than Stata does, and it's probably a cleaner way to do everything, overall. Plus R is open source (i.e. no part of any clustering methods are a black box), gives you better control over handling missing values, and has generally a more active community involved in this sort of thing. You would also have the opportunity to modify the clustering methods if you like. A couple of other alternatives would be scilab or octave (matlab-like programs), or if you're feeling adventurous, use the GSL and implement the methods yourself.

Another reason to use R over Stata for this sort of thing, is that visualization is probably important to you, and R's visual functions walk on Stata's for this sort of thing (and most other things, I believe).

your local R advocate,

Jochen Siegele wrote:

Dear all, thanks a lot for your help and comments.
We have a relatively small number of objects, around 20. Hence the raw data matrix is a 1.4 mio x 20 matrix with binary data.
By an external programme we have calculated a 20x20 distance matrix.

After importing the 20x20 distance matrix into STATA, STATA seems to interpret the imported matrix as raw data matrix and not as a distance matrix.
Hence STATA seems to calculate a further distance matrix based on the important distance matrix and does not seem be able to accept the imported matrix as a distance matrix.
Is there any possibility telling STATA that the imported matrix is a distance matrix and not a raw data matrix?
We use STATA 8.1 for our calculations.

Thanks a lot for any help.

Jochen Siegele

* For searches and help try:

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index