sql(HH.MM)中DateTime差异的总和

时间:2012-12-21 13:54:38

标签: sql sql-server sql-server-2008

我有SQL表格,其中DateDiff的格式为(hh.mm)

DateDiff   ATM
1.45       121
1.50       121
1.50       121

当我在ATM上Sum {Date}的Group by时,会显示

的结果
4.45    121

但实际的日期时差应为

5.25     121

我们如何在SQL查询中实现相同的功能

Select Sum(Cast(Convert(Varchar(10),DATEDIFF(HOUR,TicketRaisedOn,ClosedOn))+'.'+
Convert(Varchar(10),DATEDIFF(MINUTE,TicketRaisedOn,ClosedOn)%60) as Numeric(18,2)))[Down Time],ATM 
From Ticket Where Closed=1 And DATEPART(DAYOFYEAR,GETDATE())=DATEPART(DAYOFYEAR,TicketRaisedOn) 
Group BY ATM Order By [Down Time] Desc

TicketRaisedOn& ClosedOn属于DateTime

数据库是SQL Server 2008

上面的查询会打印出这样的结果(但是它的错误,因为它将数字加起来而不是日期时间格式)

Down Time       ATM
16.95           282
14.46           1811
14.20           52
14.04           936

样本数据

Select TicketRaisedOn,ClosedOn,ATM 
From Ticket 
Where ATM=282 And DATEPART(DAYOFYEAR,GETDATE())=DATEPART(DAYOFYEAR,TicketRaisedOn) 
And Closed=1

TicketRaisedOn          ClosedOn                ATM
2012-12-21 01:15:23.793 2012-12-21 15:11:59.240 282
2012-12-21 16:42:29.820 2012-12-21 18:21:30.797 282

2 个答案:

答案 0 :(得分:3)

你在做什么不起作用,因为1.5代表一个半小时,一小时50分钟。

要获得您想要的效果,请以小单位(例如秒)进行算术运算,然后将其转换为日期时间,然后将日期时间转换为最终表示形式。这是一个例子:

select right(convert(varchar(255),
                     DATEADD(second, DATEDIFF(s, TicketRaisedOn,ClosedOn), cast(0 as datetime)),
                     120),
             8)

答案 1 :(得分:3)

在格式化之前进行求和

SELECT
  ATM,
  CONVERT(VARCHAR(10), SUM(DATEDIFF(Minute, TicketRaisedOn, ClosedOn)) / 60)
  + '.' +
  RIGHT('00' + CONVERT(VARCHAR(2), SUM(DATEDIFF(Minute, TicketRaisedOn, ClosedOn)) % 60), 2)
FROM Ticket
GROUP BY ATM

Sql Fiddle:http://sqlfiddle.com/#!3/eca01/1