Question

我有一个按如下方式组织的表格：

Year |Account | Location| Measure1 |Measure2 |Measure3 

2020 |123a    |A       |100      |20%        |5
2020 |234b    |B       |75       |80%        |8
2020 |122c    |C       |80       |78%        |9

我想创建如下记录：

Year  |Account | Measure      |A    |B     |C

2020  |123a     |Measure1    |100   |      |
2020  |
2020  |234b     |Measure2    |      |80%   |
2020  |122c     |Measure3    |      |      |9

Answer 1

这个论坛不是用来解决你的编程任务的。但为了它的乐趣：

select Year, Account, 'Measure1' as Measure, Measure1 as A, null as B, null as C
from my_table
where Location = 'A'

union

select Year, Account, 'Measure2' as Measure, Measure2 as A, null as B, null as C
from my_table
where Location = 'A'

...

union

select Year, Account, 'Measure1' as Measure, null as A, Measure1 as B, null as C
from my_table
where Location = 'B'

...

Answer 2

我会推荐APPLY：

select t.year, t.account, v.*
from t cross apply
     (values ('Measure1', Measure1, null, null),
             ('Measure2', null, Measure2, null),
             ('Measure3', null, null, Measure3)
     ) v(measure, a, b, c);

这将返回除第二行之外的所有内容。不清楚你想要什么逻辑，但你可以添加：

union all
select distinct t.year, null, null, null, null, null

显式添加。

SQL 将一些列名转换为行值，将一个列名行值转换为列名

2 个答案: