There are special rules for using values for different dimensions of a segment:
Dimensions as selection criteria
Identical dimensions
(for example for senders and receivers) only need to be entered once in the sender selection criteria, since they will be inherited. You can find more information on inherited dimensions under Dimensions in Allocation.If you do not enter a dimension as selection criteria, the dimension is not included in the data records for the allocation cycle and the data is summarized for this dimension. For undefined dimensions, the system will select all values.
In iterative processing, the system will allocate back to the sender if a dimension value is the same in the sender as it is in the receiver.

The sender value is 400000 for the dimension account. The receiver values are 400000 for the dimension account and 110 and 120 for the dimension cost center. Segment processing will result in an endless loop because the receiver continues to assess/distribute back to the sender value 400000.
Sets as selection criteria
The following rules control the use of set values, if you use sets as selection criteria for your dimensions:
If you do not define a dimension, all data records for this dimension are selected; however, the initial value of the dimension is summarized in the created sender and receiver data records.
If you select all values for a dimension (for example, using a range of values or the set value All values), all data records for the dimension are selected and sender and receiver data records are created for every field value (the dimension will not be summarized).

In Customizing for the Special Purpose Ledger you can define the following fields more precisely:
- Sender/receiver
- Sender values
- Receiver tracing factor
You can determine:

For more information, refer to the activities under Allocation in the Implementation Guide (IMG) for Special Purpose Ledger.