Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Nick Cox <njcoxstata@gmail.com> |
| To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
| Subject | Re: st: RE: Check for coding of new variables using data patterns |
| Date | Tue, 22 Oct 2013 21:22:40 +0100 |
Thanks to David for the mention. -groups- was given something of a write-up in 2003 in http://www.stata-journal.com/sjpdf.html?articlenum=pr0011 but the SSC version is naturally more up-to-date. (The other programs mentioned in that article did not survive natural selection.) Nick njcoxstata@gmail.com On 22 October 2013 19:41, Radwin, David <dradwin@rti.org> wrote: > Caroline, > > How about -groups- (Nick Cox, SSC)? > > . groups var1 var2 var3 var4 var5 newvar Thirukumaran, Caroline Pinto >> Is there a way to check the coding of a newly created variable using >> data patterns of variables that have used to create the new variable? >> >> As an example, newvar is a variable that has been created based on >> values of var1-var5. The code used for creating the newvar variable is >> as follows: >> egen newvar=rsum(var1 var2 var3 var4 var5) >> >> To check that newvar has been correctly coded, it would be helpful to >> have an output like the one below: >> >> newvar var1 var2 var3 var4 var5 Frequency >> 0 0 0 0 0 0 10 >> 1 0 0 0 0 1 20 >> 1 0 1 0 0 0 40 >> 2 0 0 1 1 0 70 >> 2 0 1 1 0 0 80 >> 3 2 0 1 0 0 110 >> 3 1 0 1 0 1 120 >> 4 1 0 1 2 0 130 >> >> -collapse- gives an acceptable output (it does not give me the >> frequency count) only when var1 -var5 are binary. >> >> I am using Stata 12.1 for Windows. >> >> I get the output tabulated above from SAS using the following code: >> proc freq data=abc; >> tables newvar*var1*var2*var3*var4*var5 / list missing; >> run; * * 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/