For example, in P2P data model we connect EKPO and EKKO using foreign key relationship, but then also we mention that in data jobs while creating EKKO. So my question is why we are doing the double work here?
Also what is going to happen if I do not connect two tables that are required to be connected in the data model ?
Hello Raman,
The role of the transformations in a data job are to prep the tables that will be used in an analysis. We may have an example where P2P_EKPO may have elements that are traditionally found in EKKO, but it would make business sense to explicitly bring them into the EKPO table to drive a more efficient analysis formula down the road.
The role of a data model is to take those tables and define the schema structure. This definition is what allows aggregations and relations to occur within an asset such as an Analysis. The tables that the data model ingest do not have inherent relationships with each other, and the data model solves this by explicitly specifying the relationships.
If the tables in the data model are not connected, we will not be able to leverage the implicit joins that are frequently seen and used within analyses. So for example, if you do not connect EKKO with EKPO, the analyses will not be able to do aggregate upwards, or pull data downwards, because as far as the data model is concerned, they would not be related.
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.
