Choose How To Create Dimension (Dimension Wizard)
In this step of the wizard, you specify the type of source for the dimension's data.
Options
Star Schema: A single dimension table
Select to create a regular dimension based on a single dimension table. The depth of the dimension depends on the number of levels you select in a later step. Each level is derived from a column. Alternatively, a single datetime column can be parsed into multiple levels to create a time dimension.
When the dimension is added to a cube, the dimension table joins to the fact table. If each of the cube's dimensions is based on a single table, the cube has a star schema.
Snowflake Schema: Multiple, related dimension tables
Select to create a regular dimension based on multiple, joined dimension tables. The depth of the dimension depends on the number of levels you select in a later step. Each level is derived from a column.
When the dimension is added to a cube, only one of its dimension tables joins to the fact table. Therefore, the cube has a snowflake schema.
Parent-Child: Two related columns in a single dimension table
Select to create a parent-child dimension, which is based on two columns of the same data type. One column identifies each dimension member, and the other identifies each member's parent. For example, in an Employee table, an Employee Number column contains an identifier for each employee, and a Manager Employee Number column contains the employee number of each employee's manager. A parent-child dimension based on these two columns mirrors an organization chart.
When the dimension is added to a cube, the dimension table joins to the fact table.
Virtual Dimension: The member properties of another dimension
Select to create a virtual dimension that is based on one or more member properties in another dimension.
Mining Model: A predictable column of an OLAP mining model
Select to create a data mining dimension from a predictable column of an OLAP mining model.