Question

我尝试用HAVING子句编写以下查询，以避免编写子查询，从而使查询的可读性稍差。：

SELECT partner_id, player_id, currency_id, platform_id 
FROM account
GROUP BY partner_id, player_id, currency_id, platform_id
HAVING MAX(date) = date

我需要获取所有行，按照我在GROUP BY子句中指定的具有max date的行进行分组。但是我得到了关于

的错误

ERROR:  column "account.date" must appear in 
the GROUP BY clause or be used in an aggregate function

我在聚合函数中使用它。怎么了？

Answer 1

您可以使用window function分配行号（如果您想要排列最上面的行，则使用排名）：

SELECT * FROM (
  SELECT partner_id, player_id, currency_id, platform_id,
     ROW_NUMBER() OVER (
       PARTITION BY partner_id, player_id, currency_id, platform_id
       ORDER BY date desc) as rn
  FROM account
  ) as t
WHERE t.rn = 1

Answer 2

尝试以下查询：

SELECT partner_id, player_id, currency_id, platform_id 
FROM account
GROUP BY date, partner_id, player_id, currency_id, platform_id
HAVING MAX(date) = date

这里你忘了在group by子句中添加日期（列名）。

请告诉我，这对你有用吗？

HAVING子句有什么问题？

2 个答案: