Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: matrix glsaccum problem


From   "Richard Upward" <Richard.Upward@nottingham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: matrix glsaccum problem
Date   Sun, 16 Jun 2002 16:25:43 +0100

Dear Statalisters,

We are currently baffled by the description of 
-glsaccum- in the manual [P] pp.151-152, as there are no examples.

We are trying to create a weighted cross-product matrix of the form
X'WX, where the weighting matrix W is potentially different for each
observation in the data. W is therefore potentially n by n in dimension
(albeit with zeros everywhere except the leading diagonal), and since n
is about 1 million, we clearly cannot create W directly.

We had thought that, rather like -matrix accum-,
-matrix glsaccum- would allow us to specify a variable which contained
the weights to be used in W.  But it seems that one must specify W as a
matrix.  The manual seems to assume that W varies only across relatively
broad "groups" and is therefore quite small in dimension.

We are obviously missing something obvious, but with no concrete example
we are rather stuck.  Can anyone show us the (probably obvious)
solution?

Thanks for your time

Richard

Richard Upward
School of Economics
University of Nottingham
University Park
Nottingham NG7 2RD
Tel: +44 (0) 115 95 14735

http://www.nottingham.ac.uk/economics/staff/details/richard_upward.html
*
*   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/



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