SQL服务器：GROUP BY和DATEADD混乱

Question

我在SELECT查询与GROUP BY和DATEADD一个问题，它是这样的：（见下文样本SQL代码）

SELECT 
    'USER1' AS [user_id], 
    DATEADD(hh,-15,log_date) AS [LogDate], 
    MIN(DATEADD(hh,-15,login_time)) AS [Login], 
    MAX(DATEADD(hh,-15,logout_time)) AS [Logout], 
    DATEDIFF(MINUTE,MIN(DATEADD(hh,-15,login_time)), 
    MAX(DATEADD(hh,-15,logout_time)))/60.0 AS [TotalHrs] 
FROM 
    tblDTR 
WHERE 
    user_id = 'USER1' 
    AND log_date BETWEEN CAST('02/07/2014' AS datetime) 
        AND CAST('02/28/2014' AS datetime) 
GROUP BY 
    user_id, 
    DATEADD(hh, -15, log_date) 
ORDER BY 
    LogDate ASC 

，这是结果：

我想要做的是LogDate列必须显示真正的约会&时间（意思是，列不能减去-15），也可以在LogDate列上的结果表中看到，尽管我用GROUP BY DATEADD(hh, -15, log_date)它根本不起作用。我错过了我的SQL代码的一些功能吗？

这是我的新的SQL代码以显示LogDate没有减去15：

SELECT 
    'USER1' AS [user_id], 
    CONVERT(varchar,log_date,101) AS [LogDate], <-- I changed here but it didn't work, it give me error. 
    MIN(DATEADD(hh, -15, login_time)) AS [Login], 
    MAX(DATEADD(hh, -15, logout_time)) AS [Logout], 
    DATEDIFF(MINUTE, MIN(DATEADD(hh, -15, login_time)), 
    MAX(DATEADD(hh, -15, logout_time)))/60.0 AS [TotalHrs] 
FROM 
    tblDTR 
WHERE 
    user_id = 'USER1' 
    AND log_date BETWEEN CAST('02/07/2014' AS datetime) 
        AND CAST('02/28/2014' AS datetime) 
GROUP BY 
    user_id, 
    DATEADD(hh, -15, log_date) 
ORDER BY 
    LogDate ASC 

错误：

Msg 8120, Level 16, State 1, Line 3
Column 'tblDTR.log_date' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause.

任何替代和建议大加赞赏。谢谢！

编辑： 顺便说一句，我使用的是SQL Server 2012的

输出必须：

------------------------------------------------------------------------- 
| user_id | LogDate |  Login  |  LogOut  | TotalHrs | 
------------------------------------------------------------------------- 
| USER1 | 02/07/2014 | 2014-02-07 08:12 | 2014-02-07 19:30 | 11.28 | 
------------------------------------------------------------------------- 

Answer 1

你可以简单地包装你的选择与一个又一个，加回的15小时：

SELECT [user_id], 
    DATEADD(hh, 15 ,LogDate) AS [LogDate], 
    [Login], 
    [Logout], 
    [TotalHrs] 
FROM (
    SELECT 
    'USER1' AS [user_id], 
    DATEADD(hh,-15,log_date) AS [LogDate], 
    MIN(DATEADD(hh,-15,login_time)) AS [Login], 
    MAX(DATEADD(hh,-15,logout_time)) AS [Logout], 
    DATEDIFF(MINUTE,MIN(DATEADD(hh,-15,login_time)), 
        MAX(DATEADD(hh,-15,logout_time)))/60.0 AS [TotalHrs] 
    FROM tblDTR 
    WHERE user_id = 'USER1' 
     AND log_date 
      BETWEEN CAST('02/07/2014' AS datetime) 
       AND CAST('02/28/2014' AS datetime) 
    GROUP BY 
     user_id, 
     DATEADD(hh,-15,log_date) 
) InnerQueryAlias 
    ORDER BY LogDate ASC 

Answer 2

在查询的GROUP BY子句中，将

DATEADD(hh,-15,log_date) 

与

tblDTR.log_date