Statalist The Stata Listserver

[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: do files - general question

From   Scott Cunningham <>
Subject   st: do files - general question
Date   Tue, 26 Sep 2006 13:10:46 -0400

I've noticed some limitations when running do files, and was wondering if there was a way of getting around them. These two are, I think, related, but I'll separate them anyway.

1. When I estimate numerous different regression models, following every few with its own -estout- command, there is a point (sometimes) when the -estout- command, nor the subsequent regression commands, stop being executed within the -do- framework of execution. For instance, I might have:

. regression 1
. estimates store 1
. regression 2
. estimates store 2
. estout ...

. regression 3
. estimates store 3
. regression 4
. estimates store 4
. estout ...

and so on...

At some point, if I have too many of these repeated commands, the do file will not execute subsequent regression commands. It appears to be commented out by Stata itself, and so output simply is shown, but not executed - always preceded by an '>' in the Stata output. If I cut and paste these into a separate do file, and run them, then it executes fine. Is there a limit to how many regressions and -estout- commands one can do in a single -do- file?

2. Can one use a -do- file to execute multiple do files repeatedly? For instance, I wanted to use -sutex- to create summary stats for different datasets. So, for instance:

. set more off
. do e1
. do mergeall_e1
. sutex blah blah

. clear
> do e1
> do mergeall_e2
> #delimit ;
varlist not allowed

end of do-file

The above is a do file I had which which was supposed to do e1 then mergeall_e1, then sutex various commands, then clear, then do e1 again, then the second do, then do the same sutex a second time. But as you can see, after clearing once, I get the ">" commenting out of the subsequent commands - the same that happened above. I suspect do files cannot handle certain kinds of repeated commands, either do files or regressions combined with ado files, like estout, but is that the case? I've been getting around this by simply having multiple do files, but was just curious.

* For searches and help try:

© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index