Set Aggregation Levels for Measures

When fact tables and dimension tables are joined, you can set custom aggregation levels for a measure.

1. In Data Modeler, lock the model for editing.
2. In the Fact Tables area, click the fact table in which the measure is located.
3. Specify the aggregation rule for the new column that you want to become the level-based measure.
4. Click the column name, then click Levels.
5. In the Levels tab, for one or more hierarchies, use the slider to select the aggregation level for the measure.
   
   Description of the illustration data_mod33.gif
6. Click Done to return to the table editor.

About Setting Aggregation Levels for Measures

By default, measures are aggregated at the level of the dimension attributes that are selected in an analysis. For example, in an analysis that includes Sales Person and Revenue columns, the Revenue is aggregated at the level of a Sales Person.

To calculate ratios, you often need measures that are aggregated at a level that is different than the grain of the analysis. For example, to calculate the Revenue Percent Contribution for a Sales Person with respect to his department, you need Department Revenue at the Sales Person level in an analysis (Sales Person, Revenue, Revenue *100 / Revenue@Dept). In this example, Revenue@Dept has a custom aggregation level that is different from the default level.