Chapter 4. Multidimensional modeling: making analytics data accessible

This chapter is recommended for

In chapter 2, you saw how business questions could be described by creating cubes, dimensions, attributes, and measures, and how a schema contains those logical elements. Then, in chapter 3, you saw how to design and populate the data warehouse.

Mondrian uses the concept of a schema to map from the logical data structure used for analysis to the physical structure used in the data warehouse. A completed schema provides cubes that can be used for data analysis. In this chapter, you’ll see how to build a schema. (It’s a long chapter, because no matter how you slice it, multi-dimensional modeling is a dense topic. We suggest you read the first section, and then take a break before you proceed with section 4.2.)

This chapter describes the XML grammar of Mondrian schemas and the key XML elements and attributes. You’ll see in detail not only how to define the logical elements (cubes, dimensions, attributes, and measures) used in analytics, but also how to map them onto physical data structures (tables and columns) so that Mondrian knows how to get the data from the data mart.

4.1. A simple schema

4.2. Anatomy of a schema

4.3. Dimensions, hierarchies, and levels

4.4. Summary