Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | gsanchez@stata.com (Gustavo Sanchez) |
| To | statalist@hsphsun2.harvard.edu |
| Subject | RE: st: Mata : getting covariance matrix with optimize() |
| Date | Thu, 10 Mar 2011 10:21:00 -0600 |
Maxime <Maxime.To@ensae.fr> provided additional details on his original question about being able to get the covariance matrix when using the Mata optimize function with method "gn" combined with a weight matrix. We have not implemented variance estimation for this case, and Maxime discovered that -optimize_result_V()- returns a matrix of zeroes -- obviously not the correct variance matrix. One could argue that -optimize_result_V()- should report an error for this case, we will consider changing -optimize_reslt_V()- to do so. In the mean time, Maxime could use â??gmm- (new in Stata 11) to perform the estimation and obtain the corresponding covariance matrix. -gmm- is a programmable estimation command, with a simple syntax for specifying the moment conditions. Maxime could write to me privately in case he needs some assistance with syntax for â??gmm-. --Gustavo gsanchez@stata.com * * 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/