Question

我有一个值表，如下所示：

+--------+----+--------+
| Client | ID | Amount |
+--------+----+--------+
| A      |  1 |    100 |
| B      |  2 |     10 |
| C      |  1 |     80 |
| C      |  2 |     15 |
| D      |  2 |     20 |
| E      |  1 |     75 |
| E      |  2 |     10 |
| F      |  1 |     90 |
+--------+----+--------+

我想要做的是能够在此表上报告根据ID拆分的金额列值，所以它看起来像这样：

+--------+------+------+
| Client | ID 1 | ID 2 |
+--------+------+------+
| A      | 100  |      |
| B      |      | 10   |
| C      | 80   | 15   |
| D      |      | 20   |
| E      | 75   | 10   |
| F      | 90   |      |
+--------+------+------+

我猜关键是客户端同时拥有ID 1和2，它将出现在一行而不是多行。

任何关于如何做到这一点的想法都将受到赞赏。

Answer 1

你可以在...... construct

时使用Sum case进行条件求和

Select
Client
, ID1 = Sum(case when ID = 1 then Amount ELSE NULL end)
, ID2 = Sum(case when ID = 2 then Amount ELSE NULL end)
From MyTable
Group by Client

Answer 2

/*
sql server 2012
the code below can run at above 2000
*/
set nocount on
;with sourceData as
 (
   select * 
   from (values('A',1,100),
               ('B',2,10),
               ('C',1,80),
               ('C',2,15),
               ('D',2,20),
               ('E',1,75),
               ('E',2,10),
               ('F',1,90))as b(Client,ID,Amount)
 )
 select Client,[1] as ID1,[2] as ID2
 from sourceData as a pivot(max(Amount) for ID in([1],[2]))AS b



 /*
 Client ID1         ID2
------ ----------- -----------
A      100         NULL
B      NULL        10
C      80          15
D      NULL        20
E      75          10
F      90          NULL


 */

根据另一列拆分单列值

2 个答案: