There are published alogrithms for doing this - basically an EM procedure.
For example http://www2.sas.com/proceedings/sugi30/111-30.pdf
I know SAS has has a procedure for this. It seems to me that this problem
comes up often enough that Stata should have this capability also.
Al
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Nick Cox
Sent: Thursday, December 01, 2005 11:12 AM
To: [email protected]
Subject: st: RE: covariance matrix estimation with incomplete data
How would you want the program to do that?
Nick
[email protected]
FEIVESON, ALAN H. (modulo HTML)
I'm sure this has been asked before, but I can't find any reference to this
by searching Stata's FAQ's or by using findit - so I'll ask it again - Does
anyone know of a Stata program for maximum-likelihood (under a multivariate
normal model) estimation of a covariance matrix when incomplete data vectors
are present?
*
* 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/
