在 sql 中将所有与 group by 联合：不包含在聚合函数或 GROUP BY 中答案

【问题标题】：Union all with group by in sql: not contained in either an aggregate function or the GROUP BY在 sql 中将所有与 group by 联合：不包含在聚合函数或 GROUP BY 中
【发布时间】：2013-09-05 15:45:21
【问题描述】：

我的存储过程是这样的：

ALTER PROCEDURE [dbo].[Driverperformance] @Ecode     NVARCHAR(50), 
                                          @startdate DATETIME, 
                                          @enddate   DATETIME 
AS 
  BEGIN 
      DECLARE @date1 DATETIME = CONVERT(DATETIME, @startdate + ' 00:01:00.000', 
                                120); 
      DECLARE @date2 DATETIME = CONVERT(DATETIME, @enddate + ' 23:23:59.000', 
                                120) 
      ; 

      SELECT e.ecode, 
             Cast(q.dtime AS DATE) AS Date, 
             e.ename, 
             Count(q.ecode)        CntEcode 
      FROM   employeemaster_tbl e 
             JOIN transaction_tbl q 
               ON e.ecode = q.ecode 
      WHERE  q.ecode = @Ecode 
             AND dtime >= '' + @date1 + '' 
             AND dtime <= '' + @date2 + '' 
      UNION 
      SELECT e.ecode, 
             Cast(q.dtime AS DATE) AS Date, 
             e.ename, 
             Count(q.delecode)     DelEcode 
      FROM   employeemaster_tbl e 
             JOIN transaction_tbl q 
               ON e.ecode = q.delecode 
      WHERE  q.delecode = @Ecode 
             AND dtime >= '' + @date1 + '' 
             AND dtime <= '' + @date2 + '' 
      GROUP  BY e.ecode, 
                e.ename, 
                Cast(q.dtime AS DATE) 
      ORDER  BY Cast(q.dtime AS DATE) 
  END

执行此操作时出现错误：

选择列表中的“EmployeeMaster_tbl.Ecode”列无效因为它既不包含在聚合函数中，也不包含在 GROUP BY 子句。

【问题讨论】：

标签： sql-server

【解决方案1】：

你在UNION之前的代码段中漏掉了GROUP BY

ALTER PROCEDURE [dbo].[Driverperformance] @Ecode     NVARCHAR(50), 
                                          @startdate DATETIME, 
                                          @enddate   DATETIME 
AS 
  BEGIN 
      DECLARE @date1 DATETIME = CONVERT(DATETIME, @startdate + ' 00:01:00.000', 
                                120); 
      DECLARE @date2 DATETIME = CONVERT(DATETIME, @enddate + ' 23:23:59.000', 
                                120) 
      ; 

      SELECT e.ecode, 
             Cast(q.dtime AS DATE) AS Date, 
             e.ename, 
             Count(q.ecode)        CntEcode 
      FROM   employeemaster_tbl e 
             JOIN transaction_tbl q 
               ON e.ecode = q.ecode 
      WHERE  q.ecode = @Ecode 
             AND dtime >= '' + @date1 + '' 
             AND dtime <= '' + @date2 + ''
      GROUP  BY e.ecode, 
                e.ename, 
                Cast(q.dtime AS DATE)              
      UNION 
      SELECT e.ecode, 
             Cast(q.dtime AS DATE) AS Date, 
             e.ename, 
             Count(q.delecode)     DelEcode 
      FROM   employeemaster_tbl e 
             JOIN transaction_tbl q 
               ON e.ecode = q.delecode 
      WHERE  q.delecode = @Ecode 
             AND dtime >= '' + @date1 + '' 
             AND dtime <= '' + @date2 + '' 
      GROUP  BY e.ecode, 
                e.ename, 
                Cast(q.dtime AS DATE) 
      ORDER  BY Cast(q.dtime AS DATE) 
  END

拉吉

【讨论】：

如果我删除联合我会得到两个结果..实际上我想加入这两个结果..这就是我使用联合的原因，
除非你想要 DISTINCT 结果，否则你应该使用UNION ALL
我想得到这样的结果：
Ecode Date Ename CntEcode DelEcode E003 2013-05-05 Raheem 4 3 E003 2013-05-27 Raheem 0 7 E003 2013-05-28 Raheem 1 1