为来自不同表的两个单独操作运行多个子查询（相关）并加入一个表[关闭]

Posted 2023-02-16

【中文标题】为来自不同表的两个单独操作运行多个子查询（相关）并加入一个表[关闭]

【英文标题】：Run multiple subqueries(correlated) for the two separate operations from different tables and joining into one table [closed]

【发布时间】：2021-03-31 04:51:56

【问题描述】：

我正在尝试使用两个子查询，但找不到正确的方法，我的以下查询不是正确的方法。请帮助我了解如何使用创建列revenue 和spend 的子查询。一个子查询有效，但如何同时包含两者。同样的逻辑也可以通过joins推导出来，执行时间更短吗？

select country, location, postalcode, MAX(spend), MAX(revenue)
from (select a.*,
             (select SUM(r.revenue)
              from return r
              where r.uid = a.uid and
               r.dt >= a.dt - interval 10 minute and
               r.dt <= a.dt + interval 10 minute and
               r.cat_id in ('1200') and
               
             ) as revenue
             (select SUM(spend)
             from invest a
             where r.uid = a.uid and
              a.category = '433'
              a.cat_id in ('0', '1200') and
              a.event in ('begin')
             ) as spend
      from invest a
      where a.event in ('show1', 'show2', 'begin') and 
            a.category = '433' and
            
     ) a
group by country, location, postalcode

同样的逻辑也可以通过joins推导出来，执行时间更短？

**Invest Table**
dt                  user cat_id cat location  postal     event    spent
2020-11-01 22:12:25  1     0    A      US      X12        Show      2
2020-11-01 22:12:25  1     0    A      US      X12        Show      2 (duplicate also in table)
2020-11-01 22:12:25  1     6    A      US      X12        Mid      null
2020-11-01 22:13:20  2     0    B      UK      L23        Show      2
2020-11-01 22:15:24  2     3    B      UK      L23        End      null

**Revenue table**
dt                     user  cat_id   revenue
2020-11-01 22:14:45     1      6        null
2020-11-01 22:13:20     2      3        3

想要创建决赛桌（通过汇总每个“邮政”区域的收入）：

location   postal   spend  revenue returns
UK          X12      2       0        0
US          L23      2       3        3/2=1.5  

【问题讨论】：

请提供样本数据、期望的结果以及您想要完成的目标的清晰说明。

见meta.***.com/questions/333952/…

【参考方案1】：

可以完全删除其中一个子查询（支出）：

select country, location, postalcode,
             SUM(case when a.event= 'begin' and a.cat_id in ('0', '1200') then spend end) as spend,
             (select SUM(r.revenue)
              from return r
              where r.uid = a.uid and
               r.dt >= a.dt - interval 10 minute and
               r.dt <= a.dt + interval 10 minute and
               r.pixel_id in ('1200') and -------------why is this and ?
               
             ) as revenue
      from invest a
      where a.event in ('show1', 'show2', 'begin') and 
            a.category = '433' 
            
group by country, location, postalcode

可以加入第二个子查询：

select country, location, postalcode,
             SUM(case when a.event= 'begin' and a.cat_id in ('0', '1200') then spend end) as spend,
             b.revenue
      from invest a 
           left join                  
         (select SUM(r.revenue) revenue, dt, uid
              from return r
              where
               r.pixel_id in ('1200')
               group by dt
               
             ) b ON b.uid = a.uid and
                    b.dt >= a.dt - interval 10 minute and
                    b.dt <= a.dt + interval 10 minute 
      where a.event in ('show1', 'show2', 'begin') and 
            a.category = '433' 
            
group by country, location, postalcode, b.revenue

【讨论】：

我按照您建议的逻辑进行了尝试，return r 部分似乎存在语法问题。请参阅此You have an error in your SQL syntax; it seems the error is around: 'return r where r.uid = a.uid and r.dt >= a.dt - interval 10 minu' at line 16。问题似乎在FROM return r 部分

@SachinKumar 当然它在 WHERE 中包含 and 没有条件，就像在您的原始查询中一样。删除它或写一些布尔表达式

@SachinKumar 只有你可以调试它，很遗憾我无法重现，因为没有你的表和数据

谢谢，我已经用表格更新了问题。在 uid 上进行连接，但是当“where”子句的条件类似于where r.uid = a.uid 时，会不断弹出错误cannot resolve '`a.`' given input columns: [r.dt, r.referer, r.revenue, r.stamp]，这意味着它只查看一个表列，即r。是不是因为加入操作出错的原因，可以建议吗

@SachinKumar 哦，是的。错误显然在这里：r.dt >= a.dt - 间隔 10 分钟和 r.dt