combine only works on aggregate columns that belong to continuous views.

创建CONTINUOUS

1

2

3

CREATE CONTINUOUS VIEW v AS

SELECT g::integer, AVG(x::integer) FROM stream GROUP BY g;

CREATE CONTINUOUS VIEW

插入数据

1 2	`INSERT` `INTO` `stream (g, x)` `VALUES` `(0, 10), (0, 10), (0, 10), (0, 10), (0, 10);` `INSERT` `INTO` `stream (g, x)` `VALUES` `(1, 20);`

查询结果

1

2

3

4

5

6

pipeline=# SELECT * FROM v;

g | avg

---+---------------------

0 | 10.0000000000000000

1 | 20.0000000000000000

(2 rows)

1

2

3

4

5

pipeline=# SELECT avg(avg) FROM v;

avg

---------------------

15.0000000000000000

(1 row)

使用Combine查询

1

2

3

4

5

pipeline=# SELECT combine(avg) FROM v;

combine

---------------------

11.6666666666666667

(1 row)

直接对表表v求平均值，只会对对CQL结果进行再次求平均值。

如果拿Combine，那么会对总的6条数据来进行求平均值。

ITPub博客

pipelineDB里Combine用法