在 SQL 中加入两个表、分组和数据透视

Question

我有两个 table，我想在其中加入两个字段，根据 table 中的某些值创建组，然后执行数据透视。这是我的 tables:

table 1

id  date        Total   avail   
123 9/20/2020   10      2   
133 9/20/2020   20      3   
144 9/20/2020   10      1   
55  9/20/2020   15      5

table2

id  sku     date
123 Ax-89   9/20/2020
144 At-90   9/20/2020
133 By-25   9/20/2020
55  Bt-20   9/20/2020

期望的结果：

id  date    field   group   pivot   sku
123 9/20/2020   2   group1  avail   Ax-89
123 9/20/2020   8   group1  used    Ax-89
144 9/20/2020   1   group1  avail   At-90
144 9/20/2020   9   group1  used    At-90
133 9/20/2020   3   group2  avail   By-25
133 9/20/2020   7   group2  used    By-25
55  9/20/2020   5   group2  avail   Bt-20
55  9/20/2020   10  group2  used    Bt-20

我在做什么

SELECT table1.id table1.date, table1.total, table1.avail ,
       table2.id, table2.sku, table2.date
FROM   table1
JOIN   table2 ON
table1.date = table2.date
table1.id = table2.id
PIVOT(table1.avail) AS Pivot_table
GROUPBY table2.sku WHERE sku CONTAINS 'Ax', 'At' AS 'Group1' AND
WHERE sku CONTAINS 'By', 'Bt' AS 'Group2'

我还在研究中，欢迎任何建议。

Answer 1

您想逆轴旋转。在 SQL 服务器中我会推荐 cross apply:

select t1.id, t1.date, x.field, x.pivot, t2.sku
from table1 t1
inner join table2 t2 on t2.id = t1.id and t1.date = t2.date
cross apply (values (t1.avail, 'avail'), (t1.total - t1.avail, 'used')) as x(field, pivot)

尚不清楚列 group 需要哪种逻辑（顺便说一句，这是一个语言关键字，因此不是列名的好选择）。根据您的尝试，您似乎想根据 sku 分配它。如果是这样，请使用 case 表达式：

select t1.id, t1.date, x.field, x.pivot, t2.sku,
    case 
        when t2.sku like 'Ax%' or t2.sku like 'At%' then 'group1'
        when t2.sku like 'By%' or t2.sku like 'Bt%' then 'group2'
    end as grp
from table1 t1
inner join table2 t2 on t2.id = t1.id and t1.date = t2.date
cross apply (values (t1.avail, 'avail'), (t1.total - t1.avail, 'used')) as x(field, pivot)

在 SQL 中加入两个表、分组和数据透视

Join two tables, group and pivot in SQL

sql

sql-server

inner-join

unpivot

lateral-join