Question on calculated members

Helen999888

Hi, all experts,

As we know the calculated members are only calculated at query time which means the analysis services engine wont preaggregate any thing for calculated members across different levels of dimensions I am a bit confused then what is the point to have a calculated member in a cube especially if the cube is a large cube with over millions records of data In this case, the query including any calculated members against the cube with millions of records will be extremly slow

I am looking forward to hearing from you shortly for your kind advices and thanks a lot in advance.

With best regards,

Yours sincerely,

Jeffrey Wang


You are right that calculated members are evaluated at query time. While a cube may contain millions of records of data, a given query may be answered by using a small portion of the data. Therefore, a calculated member may be calculated pretty fast if the query space is much smaller than the entire cube.

PedroCGD

Dear Helen,

You are right, but you could use named calculations in the datasourceview at the each row nivel... you dont agregations and is better for performnace.

Calculated members is important to use in the cube, aggregations nivel...

Regards!

Helen999888

Hi, PedroCGD,

Thank you for your advices. So as you are saying, the analysis services engine will be able to locate the proportion of the cube easily whenever the MDX query is launched against the cube

Thank you.

Best regards,

PedroCGD

Avoid calculated members if you could create named calculations in datasource view. The named calculation will be a measure like other measure you select from your factTable. But of course, that if you need calculation inside aggregation, I must use calculated members...

Helen999888

Hi, PedroCGD,

Thanks a lot for your kind advices. It's been very helpfu.

With best regards,

Yours sincerely,

PedroCGD

It's a pleasure Helen.

If your answer is responded check the post as resolved, or if not tell me how can I help!

Thanks and regards!