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: Identifying complicated varlist for foreach loop |
Date | Fri, 15 Jul 2011 04:05:07 +0100 |
unab vvars: v* foreach v of local vvars { local stub = substr("'v",2,.) local stubs 'stubs' 'stub' }is one way to do it, it being understood that ' ' should really be left single quote and right single quote.
NickOn 15 Jul 2011, at 02:45, Kyle Peyton <kyle.peyton@unimelb.edu.au> wrote:
Hi all,I have a very large dataset with some variables that I suspect have matching observations. I want to test this expectation and eliminate superfluous variables.Essentially, I have two groups. In group 1, all variables have a prefix of "v". So we might have Vpen, Vshirt, Vfrog, etc. Let's assume I have 100 variables in group 1, all of them having this prefix "v"In group 2, I have approximately 250 variables. Of those 250 variables, 100 variables are the same as those in group 1, minus the prefix. The other 150 I am certain are different. So we would have pen, shirt and frog, etc. My expectation (hope) is that this is the case and then I will be able to simply drop the variables with the "v" prefix from my dataset.Is there a way to identify the SUFFIX of everything with a v PREFIX and then define that as a list? Then I'm thinking I can create a loop along these lines:Foreach x in varlist (this is the varlist of my 100 important variables, my non-prefixed pens and frogs, etc.)Assert `x' = v`x' (assuming this RHS works !) This could be the completely wrong logic, however.
* * 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/