atoti.AggregateCache

final class atoti.AggregateCache

Aggregate cache of a Cube.

Example

>>> from dataclasses import replace
>>> table = session.create_table("Example", data_types={"id": "int"})
>>> cube = session.create_cube(table)
>>> m = cube.measures

There is a default cache:

>>> cube.aggregate_cache
AggregateCache(capacity=100, measures=None)

Increasing the capacity and only caching contributors.COUNT aggregates:

>>> cube.aggregate_cache = tt.AggregateCache(
...     capacity=200,
...     measures={m["contributors.COUNT"]},
... )
>>> cube.aggregate_cache
AggregateCache(capacity=200, measures=frozenset({m['contributors.COUNT']}))

Changing back to caching all the measures:

>>> cube.aggregate_cache = replace(cube.aggregate_cache, measures=None)
>>> cube.aggregate_cache
AggregateCache(capacity=200, measures=None)

Disabling caching but keeping sharing enabled:

>>> cube.aggregate_cache = replace(cube.aggregate_cache, capacity=0)
>>> cube.aggregate_cache
AggregateCache(capacity=0, measures=None)

Disabling caching and sharing:

>>> del cube.aggregate_cache
>>> print(cube.aggregate_cache)
None

capacity: int

The capacity of the cache.

• If greater than 0, this value corresponds to the maximum amount of {location: measure} pairs that the cache can hold.

• If 0, caching is disabled but sharing stays enabled: concurrent queries will share their computed aggregates, but the aggregates will not be stored to be reused in later queries.

measures: Set[HasIdentifier[MeasureIdentifier] | MeasureIdentifier] | None = None

The measures to cache.

If None, all measures will be cached.

This should typically be the measures that are known to be the most expensive to compute.