Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Confirm existence of a variable(s) with wildcards in the variable name


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Confirm existence of a variable(s) with wildcards in the variable name
Date   Wed, 2 Feb 2011 09:13:29 +0000

I don't understand the need for circumspection here. If you try the
direct approach

. drop vari*

the -drop- won't work if there are no matching variable names. If you
are worried about this stopping a do-file or program that you want to
continue regardless, one of

. cap noi vari*

. noi cap vari*

should meet the purpose.

Mind you,

1. It would be very poor style in a public program (not that you're
proposing that) and dubious style in a private program to do this.

2. I'd be cautious about doing this even in a .do file. Depends what
you want to do, but there may be better ways to do it. Putting it in a
-drop- because it might be needed sounds as if you could bite you.
That said, the context might well make clear why this is perfectly
reasonable.

Nick

On Wed, Feb 2, 2011 at 3:05 AM, Eric Booth <ebooth@ppri.tamu.edu> wrote:
>
> Following along with the -confirm- and -unab- help files, I first wrote this code:
>
> *****************!
> sysuse auto, clear
>
> local myvarlist   z* m* r* price for foo bar
> *note:    z*, foo, and bar are fake variables*
>
>
> foreach v in `myvarlist' {
>        loc j     // this resets j where the variable does not exist
>                    // otherwise you get duplicates in your varlist
>        cap unab j: `v'
>        foreach x in `j' {
>        cap confirm variable `x', exact
>        if !_rc {
>                loc newvarlist  `newvarlist' `x'
>                }
>          }
>        }
> di in y "`newvarlist'"
> *****************!
>
> but then I realized that the "cap unab j: `v'" was doing the work of filtering out fake variables from getting into the 'newvarlist' local before the "capture confirm" loop that I usually use for building these types of varlists had a chance to do its work.
> So,  I reduced it to the following snippet.  However, relying on -capture- alone (& leaving out the -confirm- loop) feels clumsy to me -- but I can't think of a condition that would make it fail.  Others will probably have more straightforward solutions.
>
> ****snippet****
> foreach v in `myvarlist' {
>        loc j
>        cap unab j: `v'
>                loc newvarlist  `newvarlist' `j'
>        }
> ****snippet****
>

> On Feb 1, 2011, at 8:12 PM, Will Probert wrote:

>> I'm trying to unabbreviate a variable list that contains a wildcard
>> and then keep those variables, if the variable(s) exist.  From my
>> understanding, the 'if' option in STATA's 'keep' and 'drop' commands
>> only act on observations, not on variables.
>>
>> I'm trying code, unsuccessfully, along the lines of...
>>
>> capture confirm unab new: varia*
>> if(_rc==0){
>>      unab new: varia*
>>      local newlist `new'
>> }
>> keep `newlist'

*
*   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index