It’s very possible that even experienced Excel users have never been exposed to data models. This lack of familiarity with data models may well be the single greatest source of confusion for new Power BI Users.
A simple explanation is that data models join multiple tables together. Data models incorporate the data produced by business or other processes and facilitate the connection of diverse data elements to each other. Additionally, the structure and extent of data models reflect the requirements for reporting and analytics.
The Data model used most commonly in Power BI is the star-schema, in which a central “fact” table is surround by “dimension” tables. Fact tables store facts such as sales amount, number of transactions or revenue per customer. Dimension tables are used to describe and aggregate information from the fact table.
Figure 1: A Start Schema Data Model
As was briefly discussed in the introduction to data tables, dimension tables are also necessary to multiple fact tables together. In the example below, the Actual and Budget tables are joined to 2 dimension tables, the Geography and Period tables. This means that Actual sales can be compared to budgeted sales in two dimensions – time and geography. In it’s present state, comparison by other dimensions, for example Product or Segment, isn’t possible because those dimension tables are not part of the data model.
Figure 2: Using Dimension Tables to join Fact Tables
Concept 4 in this series covers One-To-Many relationships and how to join tables together.
Comments