How to Implement this complex logic in BW?

[Former Member]

Hello Experts,

Need all of your expert help to implement this below complex logic.

Multiprovider has been created above a cube which has the following sample data fromat

Question_ID	Response1	Response2	Response3
1	M1	#	#
1	M2	P1	#
1	#	#	C1
1	#	#	C2
2	M1	#	#
2	#	P1	#
2	#	P2	#
2	#	#	C1

Need all of your expert help to implement this below complex logic.

All column data should be displayed in the same line item against each unique ID (suppressing the #(Not Assigned) value) in the target object(table) as given below:

Question_ID	Response1	Response2	Response3
1	M1	P1	C1
1	M2	#	C2
2	M1	P1	C1
2	#	P2	#

To implement above scenario, we wrote start routine between DSO and Cube, but result was displayed like this format truncating multiple responses:

Question_ID	Column1	Column2	Column3
1	M1	P1	C1
2	M1	P2	C1

Any ideas anybody plz. If using start routine/expert routine, we can implement this then can someone pl help me out with how to go about it.

Thanks..

[Former Member]

hi Suman,

We compressed the cube data , but it didn't help as every response types is independently associated with Question ids( like Response1 is independent of Response2 and Resonse3 and vice versa) and also as stated before for every response types their could be multiple responses.

As per your first solution , in the respective cube or DSO, there is one custom field present(count of responses) so could you please elaborate that how to go with this if we go by creating CKFs & Conditions.

Thanks..

[Former Member]

Hi Suman/Anshu

We already tried constant selection concept in the query designer but it is of no help as all characteristics' data belongs to only one infoprovider (i.e.infocube) which is only present in the multiprovider. Also in above diagram , all are characteristics which have to be displayed and for every response types(i.e. Response1, Response2 and Response3), there could be multiple responses and all responses have to be displayed side by side according to their question ids( shown in second diagram), so we went for start routine code but there also some values are truncated(shown in the last diagram).

Thanks

[former_member182470]

Hi,

Do you have No. of Docs(0DOCUMENTS) field in the DSO and Cube? You can handle your scenario in Query designer by creating CKFs & Conditions on No. of Docs KF. This is bit complicated, but it is possible.

One more thought: Did you compress the Cube data( which is your first Table)? After compression, you can just check the data.

Regards,

Suman

[former_member182470]

Hi,

This is not a very complex requirement. You can achieve this by using Constant Selection concept in the Query Designer.

The above given link will help you.

Regards,

Suman

[anshu_lilhori]

Hi,

Follow the link:

http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/70dcd7b5-6a3d-2d10-c984-e835e3707...

Hope that helps.

Regards,

AL