如果数据不唯一，则计数

Question

我有一张商店信息表，描述了哪些商店相互关联。数据可能如下所示：

| store_id | link_num | linked_store |
|???1?  |     1    |     10       |
|???1?  |     1    |     10       |
|???1?  |     2    |     11       |
|???1?  |     3    |     12       |
|???1?  |     3    |     13       |
|???1?  |     4    |     14       |

我想检查是否有一家商店链接到同一个 link_num 的不同商店。是否有可能输出如下内容的查询？

| store_id | link_num |  count | check  |
|???1?  |     1    |    2   | same   | 
|???1?  |     2    |    1   | (null) |
|???1?  |     3    |    2   | diff   |
|???1?  |     4    |    1   | (null) |

任何帮助表示赞赏。谢谢

Answer 1

您可以count(distinct ..)为此使用：

select store_id, link_num, count(*) as count,
       case 
          when count(distinct linked_store) = 1 and count(*) > 1 then 'same' 
          when count(distinct linked_store) > 1 and count(*) > 1 then 'diff' 
       end as "check"
from the_table
group by store_id, link_num;