Dimension_Types
This topic explains how data is stored in the dimension table.
Dimension Types distinguish how data is stored in the dimension table. This page describes the different types of dimensions available in the Digital.ai Agility data mart.
Type 1
In a Type 1 dimension, each row in the table contains the latest information from the operational system.
| Key | BusinessKey | Name |
| 1 | 1 | Foo |
If the asset identified with BusinessKey 1 changed names in the operational system, the data mart loader would update the existing row in the dimension with the new name. The previous name is not retained.
| Key | BusinessKey | Name |
| 1 | 1 | New Name |
Type 2
In a Type 2 dimension, a new row is inserted into the table when data changes in the operational system.
| Key | BusinessKey | Name | Version |
| 1 | 1 | Foo | (Null) |
If the asset identified with BusinessKey 1 changed names in the operational system, the data mart loader would add a new row to this dimension with the new name. The previous row remains the same.
| Key | BusinessKey | Name | Version |
| 1 | 1 | Foo | 0 |
| 2 | 1 | New Name | 1 |
Type 3
In a Type 3 dimension, a separate column is used to represents old values.
| Key | BusinessKey | Name | Old_Name |
| 1 | 1 | Foo | (Null) |
If the asset identified with BusinessKey 1 changed names in the operational system, the data mart loader would update the existing row in the dimension with the new name. The old name is moved to a separate column and the Name column holds the current name
| Key | BusinessKey | Name | Old_Name |
| 1 | 1 | New Name | Foo |