Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Gulsum Akarsu <gulsum.akarsu@yahoo.com> |
To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
Subject | Re: st: Easy N-equation GMM syntax |
Date | Mon, 18 Feb 2013 04:05:22 -0800 (PST) |
Hi, For OLS estimation, to repeat the estimation for all the countries one by one, I have used the following loop; forvalues i = 1(1)29 { display "OLS regression for group " `i' regress y x1 x2 x3 x4 if countdum==`i' } But, first you need to generate a variable (here, countdum) that takes value 1 for the first country, 2 for the second country,...and so on. Maybe, you can also use such a loop. Regards, ----- Original Message ----- From: "Whelan, Paul" <paul.whelan07@imperial.ac.uk> To: "<statalist@hsphsun2.harvard.edu>" <statalist@hsphsun2.harvard.edu> Cc: Sent: Friday, February 15, 2013 3:49 PM Subject: st: Easy N-equation GMM syntax Hi, I would like to estimate N linear equations in a GMM framework in order to test the significance of some subset of parameters across equations. i.e., on a country by country basis individually I would do: by country: ivreg2 Y X1 X2, bw(12) kernel(bartlett) robust If N = 2, I would create individual Y's and X's for each N=1,2 and do : gmm (eq1: Y_1 - {alpha01} - {beta11}*X11 - {beta12}*X12}), /// (eq2: Y_2 - {alpha02} - {beta21}*X21 - {beta22}*X22}), /// instruments(eq1: X11 X12) /// instruments(eq2: X21 X22) /// However, if N = 20 ... what is the efficient way of doing this ? Many thanks * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/