Tuesday, May 25, 2010

Master Data Classification

Master Data assigned to Multiple Characteristics for Classification
The standard Classification Functionality for Extracting the Master Data for Material/ Customer etc is not helpful when there are multiple characteristic values assigned to Master Data Object for example Check Boxes. The Standard Solution will extract only the first available Characteristic Values only. To get all the assigned Values for any given


Characteristic required Custom built extractor. Let us see that are the steps to be followed. This example guides you to bring in Customer Data for Ship To, where each Ship To Customer is assigned to different Groups for Reporting Purpose. The solution holds well for any Master Data Object, you can achieve this by restricting Class Type.


1. Create a class for Customer Class Type 011 using T. Code CL01 and

Create the Characteristics.













The Group Detail look like this










The associated values are stored in tables CAWN and CABN








2. Assign the above created groups to Ship using T.Code XD02 .

3. These Values are stored in table AUSP. This table stores all the classification data related to any Master Data objects by Class Type.

4. Create the Data source based on this table for BW Extraction
















5. Test the Extractor using RSA3.









6. Create Info object to hold Classification Assignments as Master Data. Since a customer can be assigned more than one Group 0SHIP-TO is compounded with Group. Group and its assigned Values, Ship To are added as attributes. Ship to be added two times here as we plan to create an Infoset as compounded object does not allow for Join with other object from any source. Need to create a new Ship-To with reference to 0SHIP_TO.









7 Create Transformation and DTP for the same.











8. Create Infoset between the Master data object, Billing Cubes by linking Ship To. Choose only required Objects for reporting.












9. Built a Query on the Info Set and here is the sample result.

No comments:

Post a Comment