st: Creating a matrix from an excel table

[Diego Canales <dcanales@berkeley.edu>]

Hi everyone,

I'm working with a network spatial weights matrix brought into excel  
which basically tells me the distance among my 200 observations. It is  
not laid out in matrix format since it was built like the gwt files  
from Geoda but with actual distances.

First row is just a formatting issue for gwt, but starting from row  
two, the first column is the ID for my first variable, the second  
column is the ID of the other matching variables, and the third column  
is the distance.

Here is how it looks like.

0 40000
1	56	298.5816873
1	55	466.3318352
1	2	537.7437382
1	3	736.950696
1	58	761.8313998
1	141	820.9843873
1	57	832.6894503
1	13	893.208735
1	135	896.4010734

I am trying to figure out how to bring it into Stata so I can perform  
a multidimensional scaling and create new latitudes and longitudes for  
my observations in order to run a K-means cluster algorithm, which  
uses either euclidian distances or manhattan distances as their  
default (I want to use my own network distances).

Does anyone know how could I import my data as a matrix into Stata, or  
Mata I guess? I'm aware that I am constrained with the matrix size (I  
have Stata SE, so 11,000 matrix maximum), so I'm planning on deleting  
observations that are above a certain threshold and come up with a  
smaller matrix.

Thanks,
Diego
---------------------------------------------------------------
Diego Canales
Master of City Planning | MCP'11
University of California at Berkeley
(415) 971-2916

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/