You're correct, and I'm wrong and surprised,
that Stata allows variable names with a "�" character.
(I wrote that as the symbol for the British pound
currency, just in case it appears as something else.)
This contradicts the documented rules for variable names. I
wouldn't rely on it working in all circumstances.
However, your problem remains difficult to
diagnose without precise details of what you typed
and what Stata typed in return.
You may find it more efficient to contact Stata
tech-support with at least the following
information.
1. The results of -about- in your Stata.
2. The results of a -describe- on the
two datasets in question.
3. The precise -joinby- command you
typed and what Stata typed in return.
Nick
[email protected]
Clive Nicholas
> Indeed a variable such as "ldm?pc" is not a legal variable:
> I know that if
> I were to enter a variable with '?' in it, Stata would not
> accept it; but
> it *does* with a '�' in it, yet shows up as a '?' in the
> data editor (and
> as a '�' in the variables box). As I said before, I
> performed a -joinby-
> very recently including these variables, and it worked
> without any major
> hitches. What I do not understand is why it has not this
> time, despite
> carefully labelling the variables the same in both datasets.
>
*
* 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/
