SQL 服务器查询连接

Question

目标 - 列出每个部门的男性平均工资和女性平均工资。显示部门名称和编号。

 select AVG(salary) as 'Avg salary for men' 
   From [Enterprise].[dbo].[Employee]JOIN [Enterprise].[dbo].[Department]
   ON Employee.EMPNO = Department.deptNo
   where SEX = 'm'

还在学习连接，不确定我是否需要外连接？我如何让它按部门列出平均值？

Answer 1

我假设你打算通过 deptNo 加入。

SELECT Department.Name, AVG(salary) as 'Avg salary for men' 
FROM [Enterprise].[dbo].[Employee]
JOIN [Enterprise].[dbo].[Department]
ON Employee.deptNo= Department.deptNo
WHERE SEX = 'm'
GROUP BY Department.Name

这将显示男性每个部门的平均值。您可以对女性执行相同的操作，并使用UNION 将她们包含在查询中。或者添加到您的 GROUP BY 性别，并完全删除 WHERE 子句。

GROUP BY 的列表通常会匹配SELECT 列表中的非聚合列。在这种情况下，它只是 Department.Name。

Answer 2

好吧，首先，您似乎在JOIN 上使用了来自Employee 的错误列。也就是说，您的查询应该是这样的：

SELECT  D.deptNo,
        D.deptName,
        AVG(CASE WHEN E.SEX = 'm' THEN salary END) [Avg salary for men],
        AVG(CASE WHEN E.SEX = 'f' THEN salary END) [Avg salary for women]
FROM [Enterprise].[dbo].[Employee] E
INNER JOIN [Enterprise].[dbo].[Department] D
    ON E.deptNO = D.deptNo     -- use the right column from Employee here
GROUP BY D.deptNo,
         D.deptName

Answer 3

SELECT Employee.EMPNO, Department.Name, SEX, AVG(salary) as [Avg salary] 
FROM [Enterprise].[dbo].[Employee]JOIN [Enterprise].[dbo].[Department]
   ON Employee.EMPNO = Department.deptNo
GROUP BY Employee.EMPNO, Department.Name, SEX

这将为您获取数据，您可以从这里随意格式化。

Answer 4

SQL OUTER JOIN 子句是 SQL JOIN 子句的一种变体，它允许 SELECT 语句访问多个表。 JOIN 子句控制表的链接方式。它是 SQL FROM 子句的限定符。因此，您需要 OUTER JOIN 在您的情况下。如果需要，您可以访问link 了解更多信息。

要获得按部门平均列出的输出，只需将其添加到查询的末尾：

ORDER BY Department.deptNo

Answer 5

没有人解释过某些部门可能没有男性或女性。

SELECT  D.deptNo,
        D.deptName,
        Coalesce(AVG(CASE WHEN E.SEX = 'm' THEN salary END), 0) [Avg salary for men],
        Coalesce(AVG(CASE WHEN E.SEX = 'f' THEN salary END, 0)) [Avg salary for women]
FROM [Enterprise].[dbo].[Department] D
LEFT JOIN [Enterprise].[dbo].[Employee] E
    ON E.deptNO = D.deptNo     -- use the right column from Employee here
GROUP BY D.deptNo,
         D.deptName

Answer 6

正如所指出的，您的加入条件没有意义，您正在将员工编号加入部门编号，这将不匹配（或将，但不会实现您想要的逻辑）。

SELECT
  d.name
, AVG(e_man.salary) avg_man
, AVG(e_woman.salary) avg_woman
FROM department d
LEFT OUTER JOIN employee e_man ON e_man.deptNo = d.deptNo AND e_man.sex = 'm'
LEFT OUTER JOIN employee e_woman ON e_woman.deptNo = d.deptNo AND e_woman.sex = 'f'
GROUP BY d.name

在这里您加入两次，一次加入男性行，一次加入女性行，并根据它们计算平均值。如果没有任何男性/女性，则使用 OUTER 连接。

---

注意：我的查询将为每个部门返回一行，男性和女性的平均值为 2 列。替代解决方案是为每个部门返回 0-2 行，每个性别有单独的行，并列出每行/性别的相关薪水。

---

编辑：人们似乎非常不满意的替代子查询解决方案。根据我的经验，有时子查询是正确的答案，有时连接是正确的答案，只要不牺牲性能，您通常会使用最适合您的逻辑目标/表示的任何一个。无论如何，您正在使用的 RDBMS 的优化器通常会重写它。与往常一样，最好检查一下计划，看看发生了什么。

SELECT
  d.name
, (
    SELECT AVG(e1.salary)
    FROM employee e1
    WHERE e1.sex = 'm'
    AND e1.deptNo = d.deptNo
  ) avg_man
, (
    SELECT AVG(e2.salary)
    FROM employee e2
    WHERE e2.sex = 'f'
    AND e2.deptNo = d.deptNo
  ) avg_woman
FROM department d

Answer 7

Query1 :

SELECT Table1.*, Table2.[Roll no],Table2.Address,Table2.City
FROM (Table1 right JOIN Table2 ON Table1.[roll no] = Table2.[roll no] and Table1.[status] = Table2.[status]) 
where Table1.Status="New" or Table2.status="New"
UNION SELECT Table1.*,Table2.[roll no], table2.Address,table2.City
FROM (Table1 left JOIN Table2 ON Table1.[roll no] = Table2.[roll no] and Table1.[status] = Table2.[status])
where Table1.Status="New" or Table2.status="New";


Qury2:

SELECT Table3.*,table4.[roll no],Table4.Sports_id,Table4.Name
FROM (Table3 right JOIN Table4 ON Table3.[roll no] = Table4.[roll no] and Table3.[status] = Table4.[status]) 
where Table3.Status="New" or Table4.status="New"
UNION SELECT Table3.*,table4.[roll no],Table4.Sports_id,Table4.Name
FROM (Table3 left JOIN Table4 ON Table3.[roll no] = Table4.[roll no] and Table3.[status] = Table4.[status])
where Table3.Status="New" or Table4.status="New";


Query3:

SELECT * 
FROM (Query2 right JOIN Query3  ON Query2.[Table1.roll no] = Query3.[table3.roll no]) 
UNION SELECT *
FROM (Query2 left JOIN Query3  ON Query2.[Table1.roll no] = Query3.[table3.roll no]);


Query4:

SELECT Query4.* INTO Final_table
FROM Query4;


Query5:

UPDATE Final_Table SET Query2_Status = "New"
WHERE Query2_Status is null or Query2_status="";