Error in query(on 29th by Naren)

Question

Hi all,
I have the following query.
SELECT S.SCHOOLNAME,
CASE WHEN SUM(SUM(DISTINCT L.WORK_SPACE))/5*100/COUNT(DISTINCT L.QUESTION_ID)<50 THEN COUNT(DISTINCT L.RESPONSER_ID) END AS RSCORE,
CASE WHEN SUM(SUM(DISTINCT L.WORK_SPACE))/5*100/COUNT(DISTINCT L.QUESTION_ID)<50 THEN ROUND(SUM(SUM(DISTINCT L.WORK_SPACE))/5*100/COUNT(DISTINCT L.QUESTION_ID),2) END AS AVERAGE
FROM TQDB_LEARNER_RESPONSE L LEFT JOIN
TQDB_LEARNER_RESPONSERS LR ON LR.RESPONSER_ID=L.RESPONSER_ID LEFT JOIN
SCHOOLS S ON S.SCHOOLID=LR.SCHOOL_ID
GROUP BY S.SCHOOLNAME
I am getting the following error.
ORA-00937:not a single - group group function
Pls help me

Answer 1

Hi,
What are the nested SUMs supposed to do? "SUM (DISTINCT l_work_space)" has one value per group. What are you trying to accomplish with "(SUM (SUM (DISTINCT l_work_space) ..."?
Aside from that, Oracle apparantly can't figure out that your CASE statements correspond to the groups. So give it some help: compute all the aggregates in an in-line view and then do the CASE statements in the main query.
For example:
SELECT SCHOOLNAME,
    CASE
        WHEN WORK_SPACE_SUM / 5 * 100 / QUESTION_COUNT < 50
        THEN RESPONSER_COUNT
    END AS RSCORE,
FROM    ( -- Begin in-line view to get group values
        SELECT S.SCHOOLNAME
        ,      SUM (DISTINCT L.WORK_SPACE)     AS WORK_SPACE_SUM
        ,      COUNT (DISTINCT L.QUESTION_ID)  AS QUESTION_COUNT
        ,      COUNT(DISTINCT L.RESPONSER_ID)  AS RESPONSER_COUNT
        ) -- End in-line view to get group valuesAlso, "SUM (DISTINCT" is very suspicious. You may use it to discount multiple copies of the same row caused by the join, but it will also discount true duplicates. Analytic functions may help here.
I'm sorry that I'm so vague. If you post a small set of sample data (just enough to illustrate why you think you need nested SUMs or SUM DISTINCT) and the results you need to get from that sample data, then I can give you a specific solution.
Message was edited by:
Frank Kulash