For the record, my explanation needs some improvement:
Stata sees $ and would be prepared for a macro name
following. But no legal macro name follows, so the
$ is treated as a literal character, as are the
characters (S_DATE) following.
Nick
[email protected]
Donald Spady
> Nick and Michael
> I must blame this on poor eyesight and seeing what I
> 'expected' to see.
> Thanks for the heads up and the explanation.
Nick Cox
> Michael's code specified curly braces { } in
>
> ${S_DATE}
>
> and he meant precisely what he wrote.
>
> In context, the braces make it clear to Stata precisely what
> is (and what is not) the name of the global macro whose
> contents you want to be inserted at that point.
>
> Common or garden parentheses ( ) have no such role
> and just get taken literally in this context. Also,
> ordinary algebra in which the choice between { } and ( )
> can just be a matter of taste or convenience is no
> guide here, as they are not equivalent in Stata.
>
> Within
>
> $(S_DATE)
>
> Stata sees a $ and gears up for a global macro name
> to follow -- but then the next thing (next token, in
> Stataspeak) is a parenthesis, which cannot be part
> of a global macro name. The result is that you just get an empty
> string inserted instead, which is naturally difficult
> to see there. (Be assured, it really is there. Or it
> really isn't there. Same thing in practice, and
> consult the theologians or philosophers about the
> difference in principle.)
>
> So in this command line the substring
>
> (S_DATE)
>
> will be taken literally and no macro subsitution will
> take place. The rest you know, I think.
>
> Nick
> [email protected]
>
*
* 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/
