Detail objects are associated with dimensions and provide additional information about the dimension.
Web Intelligence XI R2 requires a one-to-one relationship between dimensions and details (this means that a detail object can have one value only for each value of its associated dimension) and does not take detail objects into account when synchronizing data. The following example illustrates why this is necessary.
Previous versions of Web Intelligence, as well as Desktop Intelligence and BusinessObjects, allow a one-to-many relationship between dimensions and details. If you migrate a report created using any of these products and the detail object contains multiple values, Web Intelligence places the #MULTIVALUE error in the detail cell.
In this example you have two data providers, and [Address] is a detail object related to [Customer]:
| Customer | Address | Revenue |
|---|---|---|
| John | London | 10000 |
| Paul | Liverpool | 15000 |
| Customer | Number of sales |
|---|---|
John |
1234 |
| Paul | 5678 |
If you create a merged Customer dimension to synchronize the data providers, and Address can have more than one value for each customer, the result is ambiguous because there is no common value around which Web Intelligence can synchronize the data.
For example, Paul might have addresses in Liverpool and London, which means that there is no unique 'Paul' row with which Web Intelligence can synchronize Paul's telephone number. Paul has a different telephone number for each address, and Web Intelligence does not know which address to associate with the telephone number:
| Customer | Address | Telephone Number |
|---|---|---|
| John | London | 1234 |
| Paul | #MULTIVALUE | 5678 |
If the relationship between Customer and Address is one-to-one, Web Intelligence can ignore Address in the synchronization. This removes the ambiguity:
| Customer | Address | Telephone Number |
|---|---|---|
| John | London | 1234 |
| Paul | Liverpool | 5678 |