Sql 如何将有重复项的组外不同项相加

我正在计算四个组中每人的平均订单数，我有以下SQL

SELECT      TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD'),
            GROUP_P,
            COUNT(CASE WHEN ORDER_TIME IS NOT null
                       THEN ORDER_ID END) AS Numerator,
            COUNT(DISTINCT
                  CASE WHEN ORDER_TIME IS NOT null
                       THEN PERSON_ID END) AS Denominator
FROM        ORDER_TAB 
WHERE       TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD')
BETWEEN     DATEADD( day, -1, current_Date) AND current_Date
GROUP BY    1, 2

我的结果是这样的

to_date     group_p    numerator   denominator
2021-06-01  A          28          20
2021-06-01  B          5           5
2021-06-01  C          1577        1025
2021-06-01  D          1010        765

因为一个

person\u id

可以在不同的组中，如果我从上表总结，由于可能存在重复，我将无法获得准确的分母值。在我下面的代码中，我删除了

分组by

，而

分子

相加相同，但

分母

越来越少

SELECT      TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD'),

            COUNT(CASE WHEN ORDER_TIME IS NOT null
                       THEN ORDER_ID END) AS Numerator,
            COUNT(DISTINCT
                  CASE WHEN ORDER_TIME IS NOT null
                       THEN PERSON_ID END) AS Denominator
FROM        ORDER_TAB 
WHERE       TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD')
BETWEEN     DATEADD( day, -1, current_Date) AND current_Date
GROUP BY    1

由于我们的报告要求，我必须以

group\u p

列显示的方式保留输出

---

从输出中求和时，使用准确的

1699

的最佳方法是什么？

下面的查询将为分母列计算一次人员id（升序中的第一组）。我使用了row_number（）窗口排序函数和常用的表表达式

with cte as
(
    SELECT      GROUP_P,
                ORDER_TIME,
                row_number()over(partition by person_id order by group_p) rnk            
    FROM        ORDER_TAB 
    WHERE       TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD')
    BETWEEN     DATEADD( day, -1, current_Date) AND current_Date
)
select    TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD'),
          GROUP_P,
          COUNT(CASE WHEN ORDER_TIME IS NOT null
                     THEN ORDER_ID END) AS Numerator,
          COUNT(CASE WHEN ORDER_TIME IS NOT null and rnk=1
                     THEN PERSON_ID END) AS Denominator
from cte

---

请提供样本数据、所需结果、所需逻辑的清晰解释以及适当的数据库标记。如果一个人在一个多组中，那么他/她应该被计算在哪个组中？第一次输出时分子值正确吗？

with cte as
(
    SELECT      GROUP_P,
                ORDER_TIME,
                row_number()over(partition by person_id order by group_p) rnk            
    FROM        ORDER_TAB 
    WHERE       TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD')
    BETWEEN     DATEADD( day, -1, current_Date) AND current_Date
)
select    TO_DATE(ORDER_TIME AT TIME ZONE 'UTC' AT TIME ZONE 'US/Pacific','YYYY-MM-DD'),
          GROUP_P,
          COUNT(CASE WHEN ORDER_TIME IS NOT null
                     THEN ORDER_ID END) AS Numerator,
          COUNT(CASE WHEN ORDER_TIME IS NOT null and rnk=1
                     THEN PERSON_ID END) AS Denominator
from cte