If you just want to get rid of the warning "<WARNING> NULL value assigned to a variable with no INDICATOR variable (2513)" then a single var is fine. But, if you want to actually use the indicator variables to determine if data is null then you need different names.
For example: SELECT Col1,Col2,Col3 INTO v1 IND vI1,v2 IND vI2,v3 IND vI3 FROM Table If all of the vars are NOT NULL then (.v1+.v2+.v3)=0. If any of the vars are NOT NULL then (.v1*.v2*.v3)=0. If all of the vars are NULL then (.v1*.v2*.v3)<>0. If any of the vars are NULL then (.v1+.v2+.v3)<>0. Regards, Stephen Markson The Pharmacy Examining Board of Canada 416.979.2431 x251 From: [email protected] [mailto:[email protected]] On Behalf Of [email protected] Sent: Wednesday, August 29, 2012 4:07 PM To: RBASE-L Mailing List Subject: [RBASE-L] - Re: INDICATOR variables I always use the same name for every indicator variable. The datatype is always integer so no conflict there. Unless you plan on checking the indicator variable by name, I know of no reason why the names should be unique. Karen In a message dated 8/29/2012 3:04:02 PM Central Daylight Time, [email protected] writes: When using a SELECT . . . INTO . . . .INDICATOR statement can just one generic INDICATOR variable be used for each value or should there be a unique INDICATOR variable for each value? For example is it OK to do this: SELECT custname,custcity,custzip INTO vname INDICATOR vind,vcity INDICATOR vind,vzip INDICATOR vind . . . . Or is this the best way: SELECT custname,custcity,custzip INTO vname INDICATOR vnameind, vcity INDICATOR vcityind,vzip INDICATOR vzipind . . . . I'm guessing unique may be better but can a generic one be used? Any and all guidance is greatly appreciated. Thanks.
