Question

I need to update the second record AFS_OH column (ForecastID =23) with the previous AFS_ToProduce value (ForecastID = 22) from the first record below. In otherwords the second record value for AFS_OH = 307923.

Also, I need to update these values for the entire table so I the ForecastID numbers will differ based on the criteria of AFS_ToProduce < 0 as you will notice in my WHERE clause below.

My script below is updating the same record value instead of the following record. Thanks for your help!

ForecastID        AFS_OH     AFS_ToProduce
22                307923    -307923
23                     0     316602

Here’s my approach:

UPDATE FC_Test
SET AFS_OH = (AFS_ToProduce * (-1))
FROM FC_Test S3
INNER JOIN
(
    SELECT S1.FORECASTID, S2.AFS_ToProduce AS AFS_OH
    FROM FC_Test S1
    LEFT OUTER JOIN
    (
        SELECT *
        FROM FC_Test
    ) S2
    ON S1.ForecastID = S2.ForecastID
)S4 ON S3.ForecastID = S4.ForecastID
WHERE AFS_ToProduce < 0

Answer 1

The following update will transfer negative value of AFS_ToProduce to next row if next row has positive AFS_ToProduce. If not, this row will be ignored.
Inner select top 1 will retrieve previous row if AFS_ToProduce is negative; if not, current row from fc_test will be skipped. This allows for gaps in ForecastID.

update fc_test
set AFS_OH = - fc2.AFS_ToProduce
from fc_test
cross apply
(
  select fc2.AFS_ToProduce
  from 
  (
    select top 1
        fc2.AFS_ToProduce
      from fc_test fc2
     where fc2.ForecastID < fc_test.ForecastID
     order by fc2.ForecastID desc
  ) fc2
  where fc2.AFS_ToProduce < 0
) fc2
where fc_test.AFS_ToProduce > 0

TEST is here on Sql Fiddle.