Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: 'xtgls' for panels that exceed matsize limit
From
Ramona Meyricke <[email protected]>
To
[email protected]
Subject
st: 'xtgls' for panels that exceed matsize limit
Date
27 Nov 2011 21:43:58 +0000
Hello, I have an unbalanced panel of 1,248,000 observations on 10,853
individuals for up to 264 months. I want to run panel regressions allowing
for cross sectional dependence using 'xtgls' and 'xtmg'.
'xtreg' works on this panel OK, but 'xtgls' and 'xtmg' do not work and
return the error: "matsize too small - should be at least 10625". I am
using STATA IC, so the max matsize I can get is 800. I tried bootstrapping
with size(800) to try and get 'xtgls' and 'xtmg' working on a bootstrapped
sample within a matsize limit of 800, but can't get this working. The
command:
bootstrap, size(800) cluster(gvkey2) idcluster(newid) : xtgls y MktRF RF,
i(newid)
returns the same error "matsize too small - should be at least 10625". (But
I can get the same command working when I draw 800 IDs and run xtgls on the
sub-sample generated by: bsample 800, cluster(gvkey2) idcluster(newID)).
Is it possible to get bootstrap working with xtgls with over 800
individuals? Or is there another way to run panel regressions controlling
for cross-sectional dependence on this panel? (I have never programmed in
MATA, so would like to stick to STATA if possible.)
Many Thanks, Ramona
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/