Show TOC

# Multi-Dimensional Exception Aggregation

In an exception aggregatoin, you can use up to five reference characteristics, so that the aggregation rule will be applied in a calculation step on the combination of all these characteristics.

Example

You want to define a calculated key figure Sales per Employee and Product and Year.

• Variant 1 without Multi-Dimenstional Aggregation

You have to model three key figures in total, with the latter two building on the preceding one:
• Sales per Employee
• Sales per Employee and Product
• Sales per Employee and Product and Year
• Variant 2 with Multi-Dimenstional Aggregation

You can define a single calculated key figure that contains all three characteristics as reference characteristics.

This serves to simplify modeling. Thanks to all exception aggregation rules being performed on single records, without the need to calculate the substeps, performance is faster.

With multi-dimensional exception aggregation, it is also possible to create new analyses, to calculate average values for the combination of characteristics for example.

Example The following example shows that multi-dimensional exception aggregation allows other analyses by nesting multiple exception aggregations. The question of which average sales all employees can achieve with all products can be answered in two ways, taking either the average over all products that the employees have sold (variant 1) or taking the sales that the employees have made on average with their products (variant 2).
Table 1: Average by Sales, Product and Person
Person Product Sales in EURO Average Sales in EURO
Tom O'Connor A 80 -
B 40 -
Total 120 60
Max Wall C 50 -
D 70 -
E 90 -
Total 210 70
Total 330
• Variant 1: 330/5 = 66
• Variant 2: 130/2 = 65
• Variant 1: Multi-dimensional exception aggregation according to calculation rule
`AVG (Sales) by Product, Person`
calculates on the individual values for sales of all products by all persons. This produces the following result: 330/5 = 66
• Variant 2: Nested exception aggregation according to calculation rule
`AVG( AVG (Sales) by Product) by Person`
calculates on the average sales per person. With this variant, the averages by products for each employee are formed first: 120/2 = 60 and 210/3 = 70. The result of the first calculation step is then used to form the average by the employees: 60+70 = 130. This produces the following result: 130/2 = 65