Date:         Thu, 23 Aug 2007 23:22:26 -0400
Reply-To:     "Howard Schreier <hs AT dc-sug DOT org>" <nospam@HOWLES.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Howard Schreier <hs AT dc-sug DOT org>" <nospam@HOWLES.COM>
Subject:      Re: DISTINGUISH INTEGER AND DECIMAL

On Thu, 23 Aug 2007 09:00:50 -0000, cat.. <cat.b41@GMAIL.COM> wrote:

>On Aug 22, 2:55 pm, gerhard.hellrie...@T-ONLINE.DE (Gerhard
>Hellriegel) wrote:
>> I don't understand that: in SAS there is only one type of numeric variable =
>> "float binary" to say it in PL1 syntax. There are no integers in SAS. So
>> searching a dot in the values makes no sense.
>
>It does, run this:
>
>data _null_;
>        format x 8.1;
>        x = 3;
>        y = index(x,'.');
>        put y =;
>run;
>
>Regards,
>
>Catherine

But try

data _null_;
        format x 8.1;
        do x = 3, -1234567890.123, 12345678901234567890;
           y = index(x,'.');
           put x=best12. y =;
           end;
run;

You should see

x=3 y=0
x=-1234567890 y=0
x=1.2345679E19 y=2

If you refine the technique by making the numeric-to-character conversion
explicitly and using the BEST32 format, such errors should be extremely rare.