Sometimes the start filter does not filter enough data. It could be that some subset classified table contains a lot of history data that is not relevant for testing. This is why DATPROF Subset offers the option to add additional filters on other tables. These filters are additional to the process filtering applied by subset logic. The additional filter will make the subset even smaller by filtering the table not only based on the predecessors, but also using the extra additional filter.  

Let's create an example classification, to get an clear understanding

  • Select the CUSTOMERS table as Start table
  • Enter the following start filter

    where last_name like 'B%'
    CODE

    This will subset 9866 of 93552 (11%) customers

  • Make sure that all tables are classified as Subset
  • Click in the upper menu bar Visualize‚ÜíProcess model... to get an overview of the process network. 

In process dependency network the tables CUSTOMERS_HISTORY, ORDERS and CONTRACTS will be subsetted right after CUSTOMERS table.

  • Right click the CONTRACTS table and click Properties...

  • Go to Additional filter tab

  • Enter the following additional filter

    where product_id in (132,145,124,128,114,125,102,147,159,147,138,134,111,117,126)
    CODE
  • Click Close to close the properties window and apply the additional filter text

Important

Using this additional filter will select all the contracts that belong to customers where the last name starts with the letter 'B' AND where the contract references certain products. In this case it will not lead to new customers, because it is an additional filter. The filter will however affect tables that are subsetted after the CONTRACTS table. So only the details of these contracts are subsetted. Remember: additional filters cannot select more data than the start filter indirectly provides.

  • Take a look at the process model and figure out which tables are also affected by this filter


If you got stuck, take a look at the following video to see how you can apply a additional filter.