Conditional Update through Semantic Grouping

The purpose of conditional updating is to update data records to a DataStore object only if a certain condition (logical expression) has been fulfilled. The data records should not be updated if this condition has not been fulfilled. To ensure that the most recent data record can always be checked, it must be processed together with all the other data records of the logical key. This can be ensured by using the semantic grouping. Here, all of the data records in a semantic group are processed together in one package.

The conditional update is programmed in the globally valid section of a routine, within the transformation. Two steps are required:

• If the condition has been fulfilled, the source value should be updated.

• If the condition has not been fulfilled, the value in the target is first initialized using the standard setting for the DataStore object. Since this initial value is not to be updated and the original value from the target is to be retained, however, the initial value must be overwritten again with the value from the A table or activation queue.

Further examples of using semantic groupings include the calculation of averages, checksums, and totals across a characteristic or characteristic group.