0

给定以下数据：

CardholderID   Source                                   DateTime
-------------- ---------------------------------------- -----------------------
3              2nd_Flr_Ramp_Out_Reader                  2010-06-30 13:58:42.410
3              2nd_Flr_Ramp_In_Reader                   2010-06-30 13:44:22.417
3              2nd_Flr_Ramp_Out_Reader                  2010-06-30 13:41:30.510
3              Lobby_To_Office_Reader                   2010-06-30 13:27:51.407
3              2nd_Flr_Ramp_Out_Reader                  2010-06-30 13:27:31.313
3              2nd_Flr_Ramp_In_Reader                   2010-06-30 13:27:23.203
3              2nd_Flr_Ramp_Out_Reader                  2010-06-29 12:03:04.413
3              2nd_Flr_Ramp_In_Reader                   2010-06-29 09:18:16.417
3              2nd_Flr_Ramp_Out_Reader                  2010-06-28 17:51:08.507
3              2nd_Flr_Ramp_In_Reader                   2010-06-28 12:52:28.403
3              2nd_Flr_Ramp_Out_Reader                  2010-06-28 12:04:36.407
3              2nd_Flr_Ramp_In_Reader                   2010-06-28 09:17:20.407
4              1St_Flr_To_Stairs_Reader                 2010-08-02 09:41:14.403
4              1St_Flr_Door_In_Reader                   2010-08-02 09:41:12.403
4              1St_Flr_Door_Out_Reader                  2010-07-30 18:24:36.400
4              1St_Flr_To_Stairs_Reader                 2010-07-30 14:09:54.403
4              1St_Flr_Door_In_Reader                   2010-07-30 14:09:48.403
4              1St_Flr_Door_Out_Reader                  2010-07-30 13:25:24.407

如何按天获取每个 cardholderId 的“In”源的最短时间和“Out”源的最长时间。

我尝试的第一件事是

select CardHolderId, min(DateTime) as EarliestSwipe, null as LatestSwipe
from EventTable
where source like '%In_Reader%'
group by cardholderid, CardHolderFirstName, CardHolderLastName,dateadd(dd, (datediff(dd, 0, DateTime)),0)
union 
select CardHolderId, null as EarliestSwipe, max(DateTime) as LatestSwipe
from EventTable
where source like '%Out_Reader%'
group by cardholderid, CardHolderFirstName, CardHolderLastName, dateadd(dd, (datediff(dd, 0, DateTime)),0)

但我需要结合这两个查询的结果。我的问题是，我是否需要对数据进行透视以获得我想要的结果，或者我是否可以保持在我提供的示例的轨道上。我将不胜感激任何帮助。

4

1 回答 1

5

试试这个。

SELECT CardholderID,
    MIN(CASE WHEN source like '%In_Reader%' THEN DateTime ELSE NULL END) as EarliestSwipe,
    MAX(CASE WHEN source like '%Out_Reader%' THEN DateTime ELSE NULL END) as LatestSwipe,
FROM EventTable
GROUP BY CardholderID

我排除了您的附加 GROUP BY 值，但我认为可以添加它们。

于 2010-08-02T18:26:44.780 回答