Rollup and having sum issue

Question

Hello friends, I am hoping you can help with me this simple query. Basically using rollup to get subtotals. But I need to include subtotals that add up to >10,000:
select a.column1
,nvl(to_char(m3.column2),DECODE(a.column1,null,null,'SubTotal')) somevalue
,c.column3 column3
,c.column4 column4
,c.column5 column5
,to_char(sum(round(m4.column5/100,2)),'$99,999,999.99') column6
from table1 a,table2 c,table3 m2,table4 m3,table5 m4
where
a.someid=c.someid
and a.someid=m2.someid
and m2.someid=m3.someID
and m3.someID=m4.someID
and m3.somedate > sysdate -7
group by (a.column1,c.column3,c.column4,c.column5),rollup (m3.column2)
having sum(round(m4.column5/100,2))>10000 -- Looks like I need to specify this before the rollup but
So I get this with the "having sum" :
123456     SubTotal     123456789     Some Company LLC     $15,630.00
But I want:
12456     123546          123456789          Some Company, LLC     $5,000.00
12456     512121          123456789          Some Company, LLC     $7,000.00
12456     346545          123456789          Some Company, LLC     $125.00
12456     843131          123456789          Some Company, LLC     $5.00
12456     SubTotal     123456789          Some Company, LLC     $15,630.00

Answer 1

Hi,
Welcome to the forum!
Whenever you have a problem, please post a little sample data (CREATE TABLE and INSERT statements, relevant columns only) from all tables involved.
Also post the results you want from that data, and an explanation of how you get those results from that data, with specific examples.
Simplify the problem as much as possible. Remove all tables and columns that play no role in this problem.
Always say which version of Oracle you're using.
See the forum FAQ {message:id=9360002}
The HAVING clause deals with groups; you're not concerned with the amount in each group, but the rolled up amount of several related groups put together.
One way to do what you want is to use the analytic SUM function to get the total. Analytic functions are applied after aggregate functions, so you can use the aggregate SUM to get a sum of sums.
Here's an example using the scott.emp table. Say we want to show employees and their salaries, with a sub-total for each employee, and the total for each department, but we only want toi include departments where the total is over 9000. Here's how:
WITH     got_total     AS
     SELECT       deptno
     ,       ename
     ,       SUM (sal)     AS total_sal
     ,       SUM (SUM (sal)) OVER (PARTITION BY deptno) / 2     -- See note
                             AS a_total_sal
     FROM       scott.emp
     GROUP BY  deptno
     ,            ROLLUP (ename)
SELECT       deptno
,       ename
,       total_sal
FROM       got_total
WHERE       a_total_sal     > 9000
ORDER BY  deptno
,            ename
;Output:
`   DEPTNO ENAME       TOTAL_SAL
        20 ADAMS            1100
        20 FORD             3000
        20 JONES            2975
        20 SCOTT            3000
        20 SMITH             800
        20                 10875
        30 ALLEN            1600
        30 BLAKE            2850
        30 JAMES             950
        30 MARTIN           1250
        30 TURNER           1500
        30 WARD             1250
        30                  9400Notice that deptno=10 is not included.
Since we're using the analytic clause "PARTITION BY deptno", the analytic SUM will be the total of all rows in the result set that have the same deptno. That include the Super-Aggregate row that we get by using ROLLUP, so the amount returned will actually be double the total salary in the department. That is, the total for deptno=20 will include the individual salaries 1100, 3000, and so on, plus the super-aggregate total of 10875, so the numebr returned by the analytic SUM function will be twice the actual total. That's why I include "/ 2" in the query above.
Edited by: Frank Kulash on May 15, 2012 10:18 AM