Regular Dimensions

Regular Dimensions

A regular dimension is a dimension that is not a virtual dimension, parent-child dimension, or data mining dimension. Unlike virtual dimensions, regular dimensions have associated aggregation data in the cubes in which they are used. Unlike parent-child dimensions, whose hierarchies are unbalanced, the hierarchies in regular dimensions are either balanced or ragged. (All dimensions are regular, virtual, parent-child, or data mining.)

A regular dimension contains a number of levels equal to the number of columns selected during its definition. These levels are usually sorted from most general to least general. Virtual dimensions are similar in this respect, but parent-child dimensions always contain a single meta data level that usually produces multiple levels in the end users' view.

Unlike a parent-child dimension, which is always based on a single table, a regular dimension can be based on multiple, joined tables. If a regular dimension is based on multiple tables, adding the dimension to a cube causes it to adopt a snowflake schema if it does not already have one. (If the cube already has a snowflake schema, it retains a snowflake schema.) If a regular dimension is based on a single table, and the dimension is added to a cube with a star schema, the cube retains a star schema.

Regular dimensions can be either shared or private. For more information about shared and private dimensions, see Shared and Private Dimensions.

You can create regular dimensions by using either the Dimension Wizard or Dimension Editor for shared regular dimensions, or Cube Editor to create private regular dimensions. For more information about the Dimension Wizard and Dimension Editor, see Dimension Wizard and Dimension Editor - Schema View.

To create a shared dimension using the Dimension Wizard