Implementing a Data Model
The implementation of a data model proceeds from the design phase described in the previous recipe. The design process and its supporting documentation clarify which entities to model, their granularity, the fact-to-dimension relationships, and the fact measures that must be developed. Additionally, the model mode (Import or DirectQuery) has already been determined, and any additional business logic to be implemented via M or DAX functions is also known. The different components of the model can now be developed, including data source connectivity, queries, relationships, measures, and metadata.
In this recipe, we walk through all the primary steps in the physical implementation of a model design. Three fact tables and their related dimensions are retrieved, relationships are created, and the core measures and metadata are added. When complete, the multi-fact data model can be exposed to business users for initial testing and feedback.