This topic provides an overview of the basic structural elements of multidimensional data. It also explains how Dundas OLAP Services uses this multidimensional data schema to retrieve and represent meta-data on a chart.
Understanding a Multidimensional Data Schema
Multidimensional data uses multiple, hierarchically-structured dimensions to classify information. By using multiple levels, we can introduce a powerful degree of interactivity into our data analysis.
The following terms are commonly used to describe multidimensional data:
Cube
A cube is the primary structure used to store and retrieve data within an OLAP database. You must have a cube of data in order to work with any OLAP database management system.
Note
Although the term "cube" implies a 3-D "square type" structure with only three axes (for example length or x-axis, width or y-axis, and height or z-axis), an "OLAP cube" is not a 3-D "square type" structure, and as such, can have an unlimited number of unique axes. In OLAP terminology, the term "cube" refers to a multidimensional collection of data.
Dimension
A dimension provides a way to organize data according to independent logical groups. This is the highest structure within the multidimensional data model. It is helpful to think of a dimension as a "category of information". For example, a dimension may include the logical groups Time, Sales Department, or Store Location.
Hierarchy
The hierarchy defines the overall structure, and relationships between, the members of a dimension. It can be thought of as a type of logical "tree" that defines how different branches of members are organized within a dimension. In most cases, there is only one hierarchy. Figure 1 shows an example of a hierarchy that was set up using the members of the Location dimension.