Dimension-Specific Aggregation Rules RRS feed

  • Question

  • Hi,

    Is it possible to define different aggregation rules depending on the dimensions and define in which order aggregations have to be done? For example  : doing a MAX() on date dimension (being at the lowest level for the other dimensions) and then a SUM() to calculate the different levels of aggregation on the other dimensions (location, material...).

    If this is possible, would the performance be acceptable on high volume of data ?

    Thanks a lot for your help,





    Tuesday, June 22, 2010 3:05 PM


  • Hi,

    You can try this:

    1)    Create a measure named [Maximum Value] with aggregateFuntion Max in your cube.

    2)    Create a calculated measure like this:

    create member measures.x as

    sum(existing [Date].[Date].[Date], [Measures].[Maximum Value])


    Hope this helps,


    Raymond Li - MSFT
    Thursday, June 24, 2010 3:52 AM