Cubing — Attributes

An attribute represents the basic abstraction of a database table column. An attribute contains an SQL expression that can be either a simple mapping to a table column or a more complex expression. These more complex expressions can combine multiple columns or attributes and can use all SQL functions including user-defined functions, if necessary.

The Design Studio hides much of the complexity of the attribute object definition. In the Design Studio, you do not need to explicitly define the expression template or parameter list of the attribute. If you want to create an attribute that directly maps to a column, you select the source column when you create the attribute in the Dimension wizard or the Dimension Properties window. If you want to create a calculated attribute, you can use the SQL Expression Builder window to create the source expression. The SQL Expression Builder provides lists of available attributes, columns, operators, functions, and constants.

When other attributes are used in the SQL expression that defines an attribute, the other attributes cannot form attribute reference loops. For example, if Attribute A references Attribute B, then Attribute B cannot reference Attribute A.

Each name, when fully qualified by a schema, must be unique from the names of all other attributes and measures in the database.

The following table describes the specific properties that define an attribute.

Table 1. Attribute properties
Property	Description
SQL expression template	SQL expression that defines the attribute. The template references columns and attributes by using a `{$$n}` notation, where `n` is an ordinal number that corresponds to the list of columns and attributes.
List of columns and attributes for SQL expression	Ordered list of all columns and attributes that compose the attribute. These columns and attributes are applied as specified in the SQL expression template.
Data type (schema, name, length, scale, nullable)	Determines the data type of the attribute. The data type is based on SQL data types, and composed by data type schema, name, length, scale, and nullable.