Glad you solved your problem.
Looking at variable labels is a good idea that didn't
occur to me.
As you imply, this would need more work
to cover variables that are interaction terms, which
I would regard as a standard application of -xi-.
Nick
[email protected]
K Jensen
> I've found that the following works for my needs (which was just to
> produce lists of categorical and non-categorical variables to document
> output for other people) by using the variable labels produced by xi:
> local xid
> local not_xid
> foreach v in `vars' {
> local label: var label `v'
> if strpos("`label'","==")>0 {
> local lbl = substr("`label'",1,strpos("`label'","==")-1)
> label var `v' "`lbl'"
> local xid `xid' `v'
> }
> else {
> local not_xid `not_xid' `v'
> }
> }
>
> local label2
> local xid2
> foreach x in `xid' {
> local label2 `label2' `:var label `x''
> }
> local xid2: list uniq label2
>
> I am aware that it wouldn't work to produce actual variable names for
> long names that had been abbreviated or for "non-standard" xi
> settings, but it solves my specific problem for the dataset I am
> working on.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
