[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Suggestion for -confirm var- option
Kaleb Michaud <firstname.lastname@example.org> writes,
> [...] debug a program, when I realized that the -confirm var- command does
> not require the variable to be spelled out fully.
> For example, suppose you have a program that looks to see if the variable
> "car" exists. If it doesn't, it creates or merges it in. But, if I have a
> variable called cartoon and not car in the data:
> . confirm var car
> gives us _rc == 0 even though the var car doesn't exist and the var cartoon
> does. Any way that the Stata powers that be could add an "exact" option
That's a good idea. In the meantime, the "right" way to confirm whether
variable "car", not abbreviated, exists is
. confirm name car
. confirm new var car
or, in a programming context,
. confirm name `possiblevar'
. confirm new var `possiblevar'
The first statement verifies that `possiblevar' could be a variable name, that
it does not violate Stata's naming conventions. The second line verifies that
a variable named `possiblevar' does not already exist.
* For searches and help try: