SQL 统计不同表中出现的次数答案

【问题标题】：SQL count number of occurrences in different tablesSQL 统计不同表中出现的次数
【发布时间】：2021-04-17 02:30:24
【问题描述】：

我的数据库中有 4 个表：传入信件、传出信件、本地信件和员工。在 "letters" 表中，有一个 register_employee 已经注册了这封信。

每个员工都有自己的唯一 ID（例如“3de9a23e-b927-4a27-a66a-c1f30b8c1464”）。

表格如下所示：

employee:
 id         first_name    last_name   age  

incoming:
 id      register_employee  summary     letter   adress   date_sent

“传出”和“本地”表看起来与“传入”表相同。

我需要计算每个员工注册了多少封信以及注册的类型。

示例结果：

employee      incoming  outgoing  local 
Bob Marley      45        33        5
John Travolta   31        10        9
George Bush     98        15        38

我尝试使用完全联接和嵌套的 Select 语句，但我得到的结果可能是错误的，因为我为每个员工获得了数百万个计数（我在 db 中没有这么多字母）。

【问题讨论】：

标签： sql postgresql count

【解决方案1】：

您可以使用OUTER JOIN 和GROUP BY，如下所示：

-- 更新

你可以使用多个子查询：

select e.id, e.firstname,
       (select count(*) from incoming i where i.register_employee  = e.id) as incoming_letter,
       (select count(*) from outgoing o where o.register_employee  = e.id) as outgoing_letter,
       (select count(*) from local l where l.register_employee  = e.id) as local_letter
  from employee e

【讨论】：

不幸的是它不起作用，我仍然得到很大的数字。所有表格的数字也相同link
显示您使用的确切查询以及结果
with distinct 和 without distinct
我找到了问题的解决方案（我发布了一个答案），但我会试试你的，因为我的解决方案似乎有很多无用的操作。我现在会尝试并告诉你它是否有效。 @大力水手

【解决方案2】：

我们计算不同类别下的字母数量并将它们加入到主员工表中

select employee.first_name , t1.incoming, t2.outgoing, t3.local from (    
    select * from employee 
    left join
    (select register_employee, count(*) as incoming from incoming_letters group by register_employee) t1 on employee.id = t1.register_employee
    left join
    (select register_employee, count(*) as outgoing from outgoing_letters group by register_employee) t2 on  employee.id = t2.register_employee
    left join
    (select register_employee, count(*) as local from local group by register_employee) t3 on employee.id = t3.register_employee ) v

【讨论】：

【解决方案3】：

SELECT incom.id,incom.name,incom.incoming,out.outgoings,loc.local
FROM (SELECT 
        employees.id             as id,
        employees.short_name     as name,
        COUNT(incomings.register_employee) as incoming
      FROM employees
      LEFT JOIN incomings ON incomings.register_employee = employees.id)
      GROUP BY employees.id, employees.short_name) as incom
LEFT JOIN (SELECT 
             employees.id  as id,
             COUNT(outgoings.register_employee) as outgoings
      FROM employees
      LEFT JOIN outgoings ON outgoings.register_employee = employees.id
      GROUP BY employees.id) as out
ON incom.id=out.id
LEFT JOIN (SELECT 
             employees.id  as id,
             COUNT(local.register_employee) as local
      FROM employees
      LEFT JOIN outgoings ON local.register_employee = employees.id
      GROUP BY employees.id) as loc
ON incom.id=loc.id;

【讨论】：