Define Semiadditive Behavior
Semiadditive measures do not uniformly aggregate across all dimensions. They are very common in many business scenarios. For instance, all the cubes based on snapshots of balances over time shows this issue. These snapshots are available in applications associated with securities, account balances, budgeting, human resources, insurance policies and claims, and many more. Therefore you may add semiadditive behavior to a cube to define an aggregation method for individual measures or members of the account type attribute. Now, if the cube holds an account dimension, then you can automatically set define semiadditive behavior based on the account type.
Steps to add semiadditive behavior –
- Firstly, open a cube in Cube Designer and choose Add Business Intelligence from the Cube menu.
- Secondly, in the Business Intelligence Wizard, select the Define semiadditive behavior option on the Choose Enhancement page.
- Thirdly, this wizard then guides you through the steps of identifying which measures have semiadditive behavior.
- So, with the exception of LastChild which is available in the standard edition, semi-additive behaviors are only available in the business intelligence or enterprise editions.
Steps to Define Semiadditive Behavior
Now, on the Define Semiadditive Behavior page of the wizard, you can select how to define semiadditivity by selecting one of the given option –
Turn off semiadditive behavior
- Firstly, it is used to remove semiadditive behavior from a cube in which semiadditive behavior was previously defined.
- This selection resets a measure to SUM if it is set to any of the given aggregation function types including By Account, Average of Children, First Child, Last Child, Last Nonempty Child, First Nonempty Child and None
- Also, this option does not change measures with a regular aggregation function: Sum, Min, Max, Count, or Distinct****Count.
- The wizard has detected the ‘Account” account dimension, which contains semiadditive members. The server will aggregate members of this dimension according to the semiadditive behavior specified for each account type.
- Further, it causes the system to set all measures from a measure group dimensioned by an Account type dimension to the By Account aggregation function and the server will aggregate members of the dimension according to the semiadditive behavior specified for each account type.
Define semiadditive behavior for individual measures
It is used to select the semiadditive behavior of each measure individually. The default setting is SUM (fully additive). Now, for each measure, you can select from the types of semiadditive functionality described in the following table.
Semiadditive function | Description |
---|---|
Average of Children | It is used for the aggregation of a member is the average of its children. |
ByAccount | It allows the system to reads the semiadditive behavior specified for the account type. |
Count | The aggregation is a count of members. |
Distinct Count | The aggregation is a count of unique members. |
First Child | The member value is evaluated as the value of its first child along the time dimension. |
FirstNonEmpty | The member value is evaluated as the value of its first child along the time dimension that contains data. |
LastChild | The member value is evaluated as the value of its last child along the time dimension. |
LastNonEmpty | The member value is evaluated as the value of its last child along the time dimension that contains data. |
Max | It is used to apply the standard maximum aggregation function. |
Min | It is used to apply the standard minimum aggregation function. |
None | No aggregation is applied. |
Sum | It is used to apply standard summation function |