Question

I need to get 1 row back with sum of some values by max date. But these 2 rows were updated by different persons, but only need to show the latest updated by person id.

My query is:

select v.code, sum(v.a_lines), sum(v.b_lines),
v.updated_by, f.period, max(v.updated_on)
from f_values v, f_file f where 
v.file_id = f.id
and v.code= '100002'
and 
to_char(f.period,'MM/DD/YYYY') = '03/31/2011'
group by v.code, v.updated_by, f.period

This returns 2 rows:

102 30  100 xxx 31-MAR-11   10-AUG-11
102 50  200 yyy 31-MAR-11   07-OCT-11

Now we need to return only 1 row. Something like this:

102   80   3000   yyy 31-Mar-11  07-Oct-11

I know that this can be achieved by removing “updated_by” from the column, but I need to have it populated, but it only needs to be the most recent updated_by row (‘yyy’). Hope my question is clear

Please let know if this is possible. Thanks in advance

Harish

Answer 1

Got it to work by using

MAX(updated_by) KEEP (DENSE_RANK FIRST ORDER BY updated_on DESC)

But this query does not work on HQL, so had to change to SQL query. If anyone can help on the HQL equivalent, that would be awesome.

Thanks