SQL Server条件交叉表

Question

使用MS-SQL我正在寻找构建条件交叉表查询的最佳方法，我猜是描述它的最好方法。

在这里简化它，我有一个CLIENTS数据集（每个都有一个唯一的ID）。每个客户端都被分配到一个TEAM。大多数客户只能分配到1个团队，但有些可以分配到2个，3个或最多5个团队。我想我的查询结果如下所示：

Client_ID Assigned_Team1 Assigned_Team2 Assigned_Team3 Assigned_Team4 Assigned_Team5 
--------- -------------- -------------- -------------- -------------- -------------- 
87496  Red Team  Null   Null   Null   Null 
74559  Red Team  Blue Team  White Team  Null   Null 
56345  Blue Team  Green Team  Null   Null   Null 
21473  Yellow Team  Blue Team  White Team  Red Team  Green Team 

Answer 1

为了得到你将要使用row_number()结果。 row_number将用于为client_id的每个团队分配一个连续值。然后使用该序列号将结果转换为列。有几种方法可以将数据行转换为列。 （注：我猜对表结构）

可以使用聚合函数CASE表达式：

select client_id, 
    max(case when seq = 1 then team end) Assigned_Team1, 
    max(case when seq = 2 then team end) Assigned_Team2, 
    max(case when seq = 3 then team end) Assigned_Team3, 
    max(case when seq = 4 then team end) Assigned_Team4, 
    max(case when seq = 5 then team end) Assigned_Team5 
from 
(
    select client_id, team, 
    row_number() over(partition by client_id order by team) seq 
    from clients 
) d 
group by client_id; 

见SQL Fiddle with Demo。

或者你也可以使用旋转功能：

select client_id, 
    [1] Assigned_Team1, 
    [2] Assigned_Team2, 
    [3] Assigned_Team3, 
    [4] Assigned_Team4, 
    [5] Assigned_Team5 
from 
(
    select client_id, team, 
    row_number() over(partition by client_id order by team) seq 
    from clients 
) d 
pivot 
(
    max(team) 
    for seq in ([1], [2], [3], [4], [5]) 
) piv; 

见SQL Fiddle with Demo