AggregationThreshold
¶
Source code in docs_src/capo_cleanrooms/types/aggregation_threshold.py
allowed_aggregate_expression_type
instance-attribute
¶
allowed_aggregate_expression_type: AllowedAggregateExpressionType
Specifies whether a query can aggregate a transformed column. This applies to the arguments of both aggregate and window functions. Valid values are:
COLUMNS_ONLY – A query can aggregate only a direct column reference, such as SUM(amount), or a constant. Clean Rooms rejects a query that transforms a column and then aggregates it, such as SUM(amount * 2) or SUM(ROUND(amount)).
ANY_EXPRESSION – A query can aggregate any expression. This includes arithmetic, such as SUM(price * quantity); a cast, such as SUM(CAST(amount AS DECIMAL)); a nested function call, such as SUM(COALESCE(amount, 0)); and a conditional, such as SUM(CASE WHEN region = 'EU' THEN amount ELSE 0 END).
identity_columns
instance-attribute
¶
identity_columns: AnalysisRuleColumnList
The identity column, such as user_id, whose distinct values Clean Rooms counts to enforce minimum aggregation thresholds. Currently, you can specify only one column, and its data type must be string, varchar, or char.
minimum_identity_count
instance-attribute
¶
minimum_identity_count: int
The minimum number of distinct identities that each query output group must represent. This threshold applies to all output columns in the table. To override this threshold for a specific column, use outputColumnThresholds.
output_column_thresholds
instance-attribute
¶
output_column_thresholds: NotRequired[OutputColumnThresholdList]
The per-column overrides of minimumIdentityCount. An output column without an override uses minimumIdentityCount.
type
instance-attribute
¶
type: AggregationThresholdType
The type of aggregation that the threshold enforces. Currently, the only supported value is COUNT_DISTINCT, which counts the distinct values in the identity column.