Precomputing dimensions during semantic model process
This documentation is for Kyvos 2026.4.x. To view documentation for other Kyvos versions, click here. |
|---|
Precomputing dimensions allow you to support a large number of dimensions and attributes in a single semantic model.
Reduce the combinations of dimensions that are pre-computed to reduce the semantic model size and the time to process the semantic model. This allows you to have more dimensions in a semantic model without exploding its size. You can choose to avoid those pre-aggregations, which are less significant for use in queries.
Use the following configuration property to define partial materialization:
kyvos.build.precompute.degree: Indicates the extent to which aggregates are precomputed in a semantic model. The value ranges from 1 to 10, where 1 indicates the lowest materialization that requires aggregate computations at runtime, while 10 indicates the highest materialization. The default value is 10.