st: "Insufficient disk space" error message running -qreg-

[Arne Mastekaasa <arne.mastekaasa@sosiologi.uio.no>]

Dear Statalisters and Stata people:
My original objective was to estimate a large model (25 explanatory 
variables) on a data set with 52,000 observations, using the censored 
median regression program -clad- by Jolliffe et al. Given the size, I was 
not very surprised that problems could arise in this case, but the problems 
persisted when I used -qreg- (without bootstrapping) in stead, reduced the 
model (in several steps) to one explanatory variable, and N to 7,500 (using 
-if- in the -qreg- command line). The error message typically appears after 
8 to 10 iterations. I do not get the error message, however, if I first 
-drop- all variables except two from the data set, even if I keep all the 
52,000 observations.
This raises several questions.
- The basic question is how Stata uses disk space in such cases. First, why 
do Stata use disk space at all in this case? The full data set and overhead 
takes 98,000 Kb of memory, leaving 208,000 Kb free.
- I have a c: and a d: disk on the computer. Stata is installed on d: with 
default data area d:\data. I have 4.7Gb free space on d:, but only 37Mb on 
c:. Does Stata use c: and not d:? If so, how can I change that?
- Why does the problem seem to depend on the total size of the data set, 
and not on the number of variables and observations to be used in the 
analysis?
I'll be very grateful for help on this.

Arne
*********************************************************
Arne Mastekaasa
Department of Sociology and Human Geography
University of Oslo
P.O.Box 1096 Blindern
N-0317 Oslo
Norway
Tel.: +47 22856757
Fax: +47 22855253
Email: arnema@sosiologi.uio.no
*********************************************************
*
*   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/