Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | daniel klein <klein.daniel.81@googlemail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: Stepwise saving ommited variable |
Date | Tue, 1 May 2012 17:09:32 +0200 |
Tomas, I will not comment on stepwise regression in general, but here is one way to get the variables omitted by -stepwise- . sysuse auto ,clear . loc allvars mpg trunk weight turn foreign . stepwise reg price `allvars' . loc varsin : coln e(b) . loc omitvars : list allvars - varsin . di "`omitvars'" Best Daniel -- Dear statalist, I am running a series of regressions on time series, in which I estimate the effect of certain national holidays (fathers day, superbowl, etc.) on beer consumption of selected beer brands. Since each beer brand has a different consumption pattern, brands differ in statistically significance of holidays. This is why I decided to use stepwise regression in order to include only the relevant holidays for each beer brand. Since I am running all my regressions in -forvalues- loop (I have approx 30 beer brands), I would like to save from each regression the holidays omitted by -stepwise- in order to be able to see, which holiday in general was the most often omitted, which holiday was the one influencing the most beer brands etc. How can I achieve this? I havent found any postestimation possibility to save variables omitted by stepwise. Thank you for you help! Tomas * * 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/