st: Large data sets

[celdjt@umich.edu]

I am working with a fairly large data base using survival anlysis and am at 
the point of doing analysis using time varying covariates. When I attempt 
to stsplit the data, the program tells me the expanded data base will 
contain over 2.3 million observations and needs 486 Mgs of memory. My 
approach has been to restart the analysis, set virtual memory on, set 
memory to 512M (full amount of my RAM), stset the data, and run the stsplit 
command. The program will start to run but invaribly stops responding. 
Before I spend money on addtional RAM, I was wondering if there was any 
suggestions or advice. Modelling requirements rule out cox's regression 
with tvc. Thanks.      
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/