Microsoft Office Tutorials and References
In Depth Information
Chapter 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Data Models ........................................... 73
The first three chapters of this topic introduced some basic Microsoft SQL Server PowerPivot
for Excel features for generating interesting reports from existing data. In our examples, you
discovered the need to model your data to make it easier to understand and manage. Nevertheless,
because all the examples were introductory, data modeling was not an issue. It is now time to
perform a deeper analysis of data modeling, discover what it is, and how to handle different
Data modeling is not a new concept for database analysts and administrators. These
technicians already know that a good data model is the foundation of a good database solution.
The PowerPivot world is not exceptional in that regard: a good data model is definitely the
foundation of a good reporting system for PowerPivot.
On the other hand, data modeling is somewhat new in the Excel world. Until the introduction
of PowerPivot, Excel users could query only one table at a time and so the very concept of
relationships was missing. PowerPivot is normally seen as a querying system, a tool that
makes it possible to extract certain information from tables. PowerPivot does much more
than this, indeed. It is a sophisticated columnar database able to store huge amounts of data.
Let us restate this: it is a real database. Because you need to work with a database, you do
need to understand how different tables are related to each other and which model is the
easiest and most effective at serving your needs. In short, you need to learn the basics of
You have already learned in previous chapters the concept of relationships, which is the
foundation of data modeling. But because we believe that relationships and data models
need to be well understood by the PowerPivot analyst, we are dedicating this entire chapter
to data modeling.