Aggregate functions at the time of measurement design in a multidimensional dataset (additive _ semi-additive and non-additive)
In the SSAS series-Implementing the first cube as well as the role playing dimension, measure formatting and the creation of calculated members create a cube from existing dimensions and fact data, while interpreting the role-playing Dimension role playing dimension, calculated members, calculated as The format of the staff and other content. In this article, the main analysis and understanding of the cube in the design process of aggregate functions, corresponding to different categories of measures according to the requirements of the different aggregate function selection will also be different.
Continue using the example created in the previous SSAS article to see two sets of measure dimension groups and different measures in biwork_firstcube.
Select Reseller Sales Amount Right-view Properties aggregatefunction aggregate functions Select Sum aggregation, well understood is a summation operation.
Select Reseller Sales Count to right-click the property, and the aggregate function selects Count, which is created automatically when cube cubes are created.
Compare the structure of the fact table factresellersales and the measures that you can choose to make when creating reseller sales and Internet sales measure groups. Data for all value types like UnitPrice, Productstandardcost, salesamount These three columns create measure groups on the right Fact reseller sales all appear, but one more Fact reseller Sales Count, including a Fact Internet sales count under Fact Internet sales, is clearly counting statistics on the fact table. Therefore, the Count aggregate function is used by default in the reseller Sales Count measure, and the SUM function is used by default for other measures.