st: Re: limits

[Christopher F Baum <baum@bc.edu>]

--On Saturday, November 4, 2000 2:34 -0500 Sue wrote:

> Hi. I have a question regarding adding variables. When I try to merge two
> big data files together I get a message "no room to add more variable".
> Is  this different from the error message "no room to add more
> observation",  which has to do with setting memory size? Thank you for
> the help. Sue
and Nick replied

The first message tells you that your data set would be too wide
to fit into memory, the second that it would be too long. Either way,
the remedy is to arrange more memory, usually by just telling Stata
to use more.

Well, yes, but even the fantastic manual will not help much if the chapter 
you should read is [r] limits. One of Stata's most maddening 
characteristics is its ironclad limitation (pre-JumboStata) to 2047 
variables. Two 'big data sets' might mean more than 2000+ variables in 
total, and in that case all the RAM in Taiwan will not help. In this case 
you need to discard variables from one or both of the datasets until the 
total number after a merge fits within this limit.

This has been a source of great frustration working with a dataset like the 
Fed's Survey of Consumer Finances, which even in one cross section contains 
5000+ variables. The workarounds needed to make these data conveniently 
available undermine efforts to convince colleagues that Stata (rather than 
SAS) would be capable of doing their work.

Kit Baum
Boston College Economics




*
*   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/