Scott, thank you for your patient, rapid, and very thorough response.
It seems as if I should definitely purchase the programming manual.
-capture- indeed solves the problem, and the output indicates that the
problem was indeed missing values for a handful of panels for the "gdp"
version of the DV. In the "absolute value" version of the DV, all
panels had at least one observation, which allowed the loop to run, even
if it didn't print out a full output for the singletons.
Jason
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Scott
Merryman
Sent: Friday, September 15, 2006 2:37 PM
To: [email protected]
Subject: st: RE: RE: RE: test for stationarity that accepts the "by"
command?
Jason,
I suspect it is because the observations do not exist for all levels of
aa_cow. I would double check this or -set tracedepth 1- and -set trace
on-
to find which cow is causing problems.
