This is example no is ok .
I want count distinct less white space.
DEFINE FILE SALES
CAMPO1/A17 = IF AREA EQ 'S' THEN CITY ELSE ' ' ;
CAMPO2/A17 = IF AREA EQ 'U' THEN CITY ELSE ' ' ;
END
TABLE FILE SALES
PRINT
CAMPO1 CAMPO2
ON TABLE HOLD AS PRUEBA FORMAT ALPHA
END
TABLE FILE PRUEBA
SUM
COMPUTE CAMPO11/I10 = CNT.DST.CAMPO1 -1;
COMPUTE CAMPO22/I10 = CNT.DST.CAMPO2 -1;
-*ON TABLE HOLD AS PRUEBA3 FORMAT ALPHA
END

Sombody can help me please

This message has been edited. Last edited by: Kerry,

Hi,

you're on the right track.

You're problem is with 2 × CNT.DST.
We've had a lot of trouble with it to.

The seems to be a big difference between:

SUM COMPUTE CD1/D12 = CNT.DST.FIELD_ONE;
    COMPUTE CD2/D12 = CNT.DST.FIELD_TWO;
BY  ...

and

SUM CNT.DST.FIELD_ONE
    CNT.DST.FIELD_TWO
BY  ...

The first one is handled by WF.
The second is passed to the SQL.

( [!] This difference doesn't occur with the CAR master sample. Only with out real masters. )


First, test wether the CNT.DST work. Make sure this does want you expect.

If it does, then try to exclude ' ' with the -1.

You could help me , the first count ok but second count no ok .
Sorry my english no is very good.

I need count good in the fields

thanks for everything