Hierarchies
The Hierarchies tab displays the data hierarchies automatically identified during semantic layer creation. Hierarchies define structured relationships within your data (e.g. Country → Region → City) and help the model understand how data dimensions relate to each other.
How Hierarchies Are Generated
Hierarchies are generated automatically when you run Create Semantic Model or Update Semantic Model in the Operations tab. The system analyzes the connected data sources and identifies cross-dimensional relationships.
Managing Hierarchies
Deactivating a Hierarchy
If a generated hierarchy is incorrect or not relevant to your use case, you can deactivate it. Deactivated hierarchies are excluded from the model's semantic understanding.
Adding Hierarchies
You can manually define additional hierarchies by linking attribute associations. This is useful for domain-specific relationships that the automatic analysis may not detect.
Locking Hierarchies
Lock a hierarchy to protect it from being overwritten during future semantic layer updates.
Best Practices
- Review the automatically generated hierarchies after creating or updating the semantic layer.
- Deactivate hierarchies that do not accurately represent your data.
- Lock manually created hierarchies that are essential for correct analysis.
- Do not deactivate hierarchies unless you have verified they are incorrect — removing valid hierarchies reduces analytical precision.