SQL SELECT查询

Question

假设我有一个带有三列的SQL表"Company":"department_id","employee","job".像这样的东西:

DEPARTAMENT_ID | EMPLOYEE | JOB
--------------------------------------
1              | Mark     | President
1              | Robert   | Marketing Manager
1              | Rose     | Administration Assitant
2              | Anna     | Programmer
2              | Michael  | Programmer
2              | Celia    | Sales Manager
3              | Jhon     | Sales Manager
3              | Donna    | Programmer
3              | David    | Marketing Manager

我想编写一个返回部门ID的查询,其中至少有50%的工作是相同的.

我的例子中我需要的结果就是:

DEPARTAMENT_ID |
--------------------------------------
2              |

我该如何编写这个SQL查询？我想我尝试了各种各样的东西,但我不明白:(.

Answer 1

这有点棘手。您需要将某个部门中某项工作的总人数与总人数进行比较。因此，一种方法使用两个聚合：

select department_id
from (select department_id, count(*) as numemp
      from t
      group by department_id
     ) d join
     (select department_id, max(numemp) as numemp
      from (select department_id, job, count(*) as numemp
            from t
            group by department_id, job
           ) d
     group by department_id
    ) dj
    on d.numemp <= 2 * dj.numemp;

如果您的一个部门恰好分为两份工作，那么您可能会得到重复的信息。在这种情况下，请使用select distinct.