DS_HLL_COUNT_DISTINCT (Lakehouse v2)

使用 DataSketches HLL 算法返回近似去重计数。比 APPROX_COUNT_DISTINCT 更精确。

Analyze 语法

func.ds_hll_count_distinct(get_column(table, 'user_id'))

Analyze 示例

func.ds_hll_count_distinct(get_column(table, 'user_id'))

┌───────┐
│ 10042 │
└───────┘

SQL 语法

DS_HLL_COUNT_DISTINCT(<expr>)

SQL 示例

SELECT DS_HLL_COUNT_DISTINCT(user_id) FROM page_views;

┌───────┐
│ 10042 │
└───────┘