Re: st: if command

[David Kantor <dkantor@jhu.edu>]

At 03:26 PM 2/12/2004 -0500, Danielle Ferry wrote:

> Can anyone please explain when the if command requires quotes around the 
> expression and when it does not?
>
> For example, I've found the following seemingly quirky behavior:
>
> 1) if "`1'"=="aRate" | "`1'"=="bRate" | "`1'"=="pRate" --> requires quotes
> 2) if `1'==aRate --> does not require quotes
> 3) if "`1'"=="apRatio" --> requires quotes
As far as I know, if you are comparing to a string constant, you need quotes.

Example 2 probably ought to have quotes if it is meant to test a string 
value.  It would, however, pass as a legitimate expression if...
 aRate happens to be a variable name and...
   it is numeric and so is `1'  (that is the contents of the local 1), or
   it is string, and `1' contains enclosing quotes (I haven't tried this), or
   it is either numeric or string, and `i' is also a variable name of a 
compatible type.
But in these cases, you are implicitly testing observation 1, which you 
might not have intended.

Example 2 would also pass if aRate is a scalar and `1' is numeric (and if 
`1' is a variable name, then you are testing observation 1).

In any case, while Example 2 might sometimes pass, it is probably not doing 
what you intended.

I hope this helps.
-- David

David Kantor
Institute for Policy Studies
Johns Hopkins University
dkantor@jhu.edu
410-516-5404

*
*   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/