Sql &引用；“工会”；超过；“按多维数据集分组”；

我正在使用工会操作员打印部门id、工作id、平均工资：

• 每个部门，每个部门，每个工作
• 每个部门，不管工作
• 每项工作，不分部门
• 整张桌子

而这可以通过“按多维数据集分组”轻松实现 按多维数据集分组（a，b）：（a，b）（a）（b）（，我打算使用union操作符来执行此操作：

SELECT e.department_id, e.job_id, ROUND(AVG(e.salary))
FROM employees e
GROUP BY e.department_id, e.job_id
UNION
SELECT e.department_id, e.job_id, ROUND(AVG(e.salary))
FROM employees e;

---

在第二个联盟中，即“每个部门，不管工作如何”我该如何分组？

我相信您正在寻找这样的东西：

SELECT 'JOB PER DEPARTMENT' AS GROUP,
       A.DEPARTMENT_ID,
       B.JOB_ID,
       AVG(B.SALARY)
  FROM DEPARTMENTS A
  JOIN EMPLOYEES B
    ON A.DEPARTMENT_ID = B.DEPARTMENT_ID
 GROUP BY A.DEPARTMENT_ID, B.JOB_ID
UNION
SELECT 'DEPARTMENT' AS GROUP,
       A.DEPARTMENT_ID,
       'N/A' AS JOB_ID,
       AVG(B.SALARY)
  FROM DEPARTMENTS A
  JOIN EMPLOYEES B
    ON A.DEPARTMENT_ID = B.DEPARTMENT_ID
 GROUP BY A.DEPARTMENT_ID
UNION
SELECT 'JOB' AS GROUP,
       'N/A' AS DEPARTMENT_ID, 
       B.JOB_ID, 
       AVG(B.SALARY)
  FROM DEPARTMENT A
  JOIN EMPLOYEES B
    ON A.DEPARTMENT_ID = B.DEPARTMENT_ID
 GROUP BY A.JOB_ID

---

将第二个并集中的作业id替换为伪值或仅为null。然后仅按部门id分组