Hello,

I am looking for a keyword that will let me grab the value of the next row to use in a calculation in the current row. There is already a keyword for the last row but I can't seem to find it after looking through documentation. We would like to avoid using hold files at all.

example:

TABLE FILE CAR
SUM
DEALER_COST
COMPUTE TEST/D12 = DEALER_COST + LAST DEALEAR_COST; AS TEST
-*COMPUTE TEST2/D12 = NEXT DEALER_COST - DEALER_COST; AS TEST2
BY CAR
END

Please let me know if this is something possible in version 8.

Thank you,

Michael

This message has been edited. Last edited by: <Emily McAllister>,

As far as I know there is no NEXT row. But using the LAST should be possible, it's just a question of processing where you should only have to "invert" your logic.

It is a question of "did I skiped the fence already and have to look behind or will I skip and look after". According to me, it's easier to know what has happened then knowing what is going to happen...

So I'm pretty sure that is feasible with LAST.

There is no NEXT logic. Use of NEXT or LAST logic implies that a time component is either implied or is an explicit part of the order of the data. Just reverse that so the most recent rows are first then you can use LAST. I do it all the time.

Give this a shot.

TABLE FILE CAR
SUM CAR AS 'NXTCAR'
COMPUTE LSTCAR/A16=LAST CAR;
BY COUNTRY NOPRINT
BY CAR NOPRINT
ON TABLE SET HOLDLIST PRINTONLY
ON TABLE HOLD AS NXTCAR FORMAT FOCUS INDEX LSTCAR
END
-RUN
JOIN CAR IN CAR TO LSTCAR IN NXTCAR
-RUN
TABLE FILE CAR
SUM SALES NXTCAR
BY COUNTRY
BY CAR
END