sql - 在窗口函数中计算运行总和

Question

我在 Redshift(使用 Postgres 8)中遇到这个运行总和问题:

select extract(month from registration_time) as month
 , extract(week from registration_time)%4+1 as week
 , extract(day from registration_time) as day
 , count(*) as count_of_users_registered
 , sum(count(*)) over (ORDER BY (1,2,3))
from loyalty.v_user
group by 1,2,3
order by 1,2,3
;

我得到的错误是:

ERROR: 42601: Aggregate window functions with an ORDER BY clause require a frame clause

Answer 1

您可以在同一查询级别上对聚合函数的结果运行窗口函数。在这种情况下使用子查询要简单得多:

SELECT *, sum(count_registered_users) OVER (ORDER BY month, week, day) AS running_sum
FROM  (
   SELECT extract(month FROM registration_time)::int     AS month
        , extract(week  FROM registration_time)::int%4+1 AS week
        , extract(day   FROM registration_time)::int     AS day
        , count(*) AS count_registered_users
   FROM   loyalty.v_user
   GROUP  BY 1, 2, 3
   ORDER  BY 1, 2, 3
   ) sub;

我还修复了表达式计算 week 的语法。 extract() 返回 double ，但模运算符% 不接受 double 数字。我将所有三个都转换为 integer 。

赞@a_horse commented ，您不能在窗口函数的 ORDER BY 子句中使用位置引用(与查询的 ORDER BY 子句不同)。

但是，您不能在此查询中使用 over (order by registration_time)，因为您是按 month、week、日。 registration_time 既未聚合，也未按要求包含在 GROUP BY 子句中。在查询评估的那个阶段，您无法再访问该列。

您可以重复 ORDER BY 子句中前三个 SELECT 项的表达式以使其工作:

SELECT extract(month FROM registration_time)::int     AS month
     , extract(week  FROM registration_time)::int%4+1 AS week
     , extract(day   FROM registration_time)::int     AS day
     , count(*) AS count_registered_users
     , sum(count(*)) OVER (ORDER BY 
              extract(month FROM registration_time)::int
            , extract(week  FROM registration_time)::int%4+1
            , extract(day   FROM registration_time)::int) AS running_sum
FROM   loyalty.v_user
GROUP  BY 1, 2, 3
ORDER  BY 1, 2, 3;

但这似乎相当嘈杂。 (不过性能会很好。)

旁白:我确实想知道 week%4+1 背后的目的……整个查询可能更简单。

相关:

• Get the distinct sum of a joined table column
• PostgreSQL: running count of rows for a query 'by minute'