on 12-12-2013 10:23 AM
Dear all,
BO Version : XI 4.0
I have a big problem of aggregation in WebI and I don't understand.
To simply, I have 1 dimension DATE and 3 indicators ( type sum ) QTY_1, QTY_2 , AMOUNT
In the request analysis window of WebI, I see the first problem :
1) There is a distinct on all the result of my request and I never ask it. Where can I change it ?
2) Per each date, the request repeats QTY_1 and QTY_2 per AMOUNT. Why not, if it is only for the display of the request select but, in my report, the aggregation sum is totally wrong because :
I need the distinct QTY_1 and QTY_2 and the sum( AMOUNT)
Why do I need to do it ??
In my understanding, for each date, BO have to do the correct aggregation sum by itself !? Else I don't understand the interest to use indicators
Thank you to help me.
Best regards,
Mike
In this case what you can do is count the dates and divide this to all your measures. like
=[QTY1]/Count([DATE];All)
=[QTY2]/Count([DATE];All)
=[AMOUNT]/Count([DATE];All)
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
First of all, what do you mean by "there is a distinct on all the result"? What kind of data do you want to see in your report? If it is possible, please attach a screen shot.
Secondly, there could be so many reasons you are getting wrong results in your report. You might be getting partial results, your universe might have problems etc.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Zahid,
I mean that the query start by SELECT DISTINCT but I have changed it to replace the DISTINCT_VALUES with a group by in the file parameters of my universe.
And for the moment, I search to aggregate correctly the datas with the group by condition and some additionnal dimensions if needed.
Change the aggreagtion at universe to be sum If not done
QTY_1, QTY_2 , AMOUNT
And in the SQL Provide sum ( QTY_1) , Sum(QTY_2) , Sum (Amount) .
Check if your date have timestamp attached to date , If there is timestamp then they will be counted as Separate rows :
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
86 | |
10 | |
10 | |
9 | |
6 | |
6 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.