Data marts from several fact tables
Combining elements from different fact tables
Data marts can include elements from different fact tables, enabling flexible data combinations.
Element handling:
- Common elements from different tables: Displayed once, linked to all source tables
- Different elements: All included separately
- Automatic "fact" field: Added for UNION merging to identify the source fact table
SQL script generation
Requirements:
- All selected elements must originate from a single database connection
- Multiple connections cause SQL script generation error
UNION merging:
- Combines rows from different fact tables
- The "fact" field identifies the source fact table
- Fields defined in the source fact table: populated with values
- Fields not defined in the source table: populated with NULL
- Both data marts without grouping and data marts with grouping are supported
JOIN merging (aggregated only):
- Combines based on common dimensions
- The "fact" field is not added
- Creates all combinations of dimension values
- Does not support the addition of facts
- Measure values are calculated for each combination of dimensions
- NULL if measure values for a dimension combination cannot be calculated
Configuration rules
Enabling multi-table functionality:
- The merge type field is activated when there are ≥2 fact tables
- Automatically deactivated if the data mart contains <2 fact tables
Cloning behavior:
- The new data mart inherits the merge type from the original clone
- Can be changed during or after creation
Verification and configuration
- The same verification process as for single-table data marts
- The same view management capabilities
- Filters are applied to all merged tables