Below are the snippets of a Vendor performance report, being implemented in the system.
Here the join is between Purchase data (Cube) and Purchase delivery (DSO).
Data model of the cube (ZPUR_C03)
Data model of the DSO (ZPUR_O04)
The data in the infoproviders and in the infoset is as shown below
PO line item data
PO delivery data / Goods receipt data
As is apparent, the PO quantity from ZPUR_C03 is double the correct value.
The correct value here would be 4, but the value shown is 8.
The Concatenated referen is the reference characteristic and there are two distinct characteristic values for this characteristic.
Hence, when average exception aggregation is used, with reference char as long char(60), then the key figure PO quantity’s value = 8/2 = 4.
Design of the reference characteristic
1. It should be a perfect fit with other components of the infoset –
The key of the master data table should be exactly equal to the key of the DSO in the infoset which contains our reference characteristics.
As per this the key of our infoobject = key of our DSO (i.e. Purchase document number, item number, schedule line number, material document number and material document item).
Below are the screenshots of the build of this infoobject.
The reference characteristic need not be navigational attribute as mentioned earlier.
Master data that would be created as below.
Join in the infoset.
(3.x query designer has been used for example purpose).
In 7.x query designer key figure property aggregation allows you to define the exception aggregation. You may have to create a new selection/CKF for the key figure, if the overriding of the key figure aggregation property defined in modelling is not possible.
In 3.x you have to create a calculated key figure, with the properties set as shown below.
In BI 7.x
Observe the properties of the calculated key figure.
The PO quantity here is the value without exception aggregation and the PO quantity calculated is the value after the exception aggregation is applied.