Dear David,
thank you very much for your reply!,
but I have still difficulties:
I looked up -glsaccum- in the manual, and indeed
I want to do a 'GLS-Style estimation' namely:
matrix B = inv((z*inv(z'z)*z'x)'*(inv(sigma) #
In)*x)*(z*inv(z'z)*z'x)'*(inv(sigma) # In)*y
since x,y and z are larger than matsize I used
matrix accum to generate
ZZ=z'z
ZX=z'x
and now I have:
matrix B = inv((z*inv(ZZ)*ZX)'*(inv(sigma) #
In)*x)*(z*inv(ZZ)*ZX)'*(inv(sigma) # In)*y
but still I can not generate parts of that, namely
x,y,z and inv(sigma) # In (!)
how can -glsaccum- help me I do not have
something like X'BX
thank you very much in advance
Martin Halla
-------------------------------------------
Martin HALLA
Department of Economics
Johannes Kepler University Linz
Altenbergerstr. 69
A-4040 Linz, Austria
[email protected]
Tel.: +43-732-2468-8219
Fax: +43-732-2468-2-8219
http://www.econ.jku.at/
-------------------------------------------
>>> [email protected] 01/31/05 9:09 pm >>>
Martin Halla <[email protected]> asked how he could trasform some
data
when the matrix created by
. matrix w = inv(sigma) # I(2857)
is larger than matsize.
-glsaccum- will probably do what he wants.
-David
[email protected]
*
* 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/
*
* 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/