合并多行以获得单行答案

【问题标题】：Combining multiple rows to get a single one合并多行以获得单行
【发布时间】：2021-08-27 17:12:56
【问题描述】：

源表：

----------------------------------------
| Employee Name   | department | Emp Id |
----------------------------------------
| Sam             | Sales      | 101    |
----------------------------------------
| Sam             | Finance    | 101    |
----------------------------------------
| Dirk            | marketing  | 102    |
----------------------------------------
| Dirk            | Research   | 102    |
----------------------------------------

需要输出：

------------------------------------------------------
| Employee Name   | Emp Id | department1 | department2|
------------------------------------------------------
| Sam             | 101    | Sales       | Finance    |
------------------------------------------------------
| Dirk            | 102    | marketing   | Research   |
------------------------------------------------------

您能帮我解决一下我应该使用什么函数或查询来获得上述输出吗？

【问题讨论】：

标签： sql oracle oracle11g

【解决方案1】：

有两种已知的数据透视技术

通过使用条件聚合

SELECT EmployeeName, EmpId,
       MAX(DECODE(rn,1,Department)) AS Department_1,
       MAX(DECODE(rn,2,Department)) AS Department_2
  FROM (
        SELECT t.*, 
               ROW_NUMBER() OVER 
               (PARTITION BY EmpId,EmployeeName ORDER BY Department) AS rn
          FROM t
       )
GROUP BY EmployeeName, EmpId

2.通过使用PIVOT子句

SELECT *
  FROM (
        SELECT t.*, 
               ROW_NUMBER() OVER 
               (PARTITION BY EmpId,EmployeeName ORDER BY Department) AS rn
          FROM t
       )
 PIVOT
 (
  MAX(Department) AS Department FOR rn IN (1,2)
 )

对于这两种情况都需要枚举旋转的部门列。为此，ROW_NUMBER() 解析函数非常适合

Demo

【讨论】：

非常感谢@Barbaros Özhan 的解决方案

【解决方案2】：

如果只有两个值，可以使用max()和min()：

select name, empid,
       min(department) as department1,
       nullif(max(department), min(department)) as department2
from t
group by name, empid;

【讨论】：