子表中的SQL计数

我想计算子表中的一些列。我的表格结构如下：

人

+---+----------+--------+
| Pid | Name   |Surname |
+---+----------+--------+
| 1   | Per A  | D      |
| 2   | Per B  | E      |
| 3   | Per C  | F      
+----+---------+--------+

+---+---------+-------------------+------------+-----+
| Cid | CName          | School   | Sex        | Pid |
+---+---------+-------------------+------------+-----+
| 1   | John           | High     |    Man     | 1   |
| 2   | Alice          | Primary  |    Woman   | 2   |
| 3   | Mel            | High     |    Man     | 3   |
| 4   | Angelina       | High     |    Woman   | 2   |
+----+---------+------------------+------------+-----+

儿童

+---+----------+--------+
| Pid | Name   |Surname |
+---+----------+--------+
| 1   | Per A  | D      |
| 2   | Per B  | E      |
| 3   | Per C  | F      
+----+---------+--------+

+---+---------+-------------------+------------+-----+
| Cid | CName          | School   | Sex        | Pid |
+---+---------+-------------------+------------+-----+
| 1   | John           | High     |    Man     | 1   |
| 2   | Alice          | Primary  |    Woman   | 2   |
| 3   | Mel            | High     |    Man     | 3   |
| 4   | Angelina       | High     |    Woman   | 2   |
+----+---------+------------------+------------+-----+

所以我想输出

+---+----------+------+---------+--------+---+--------------+
| Pid| PerName | High | Primary | Woman  | Man | ChildCount | 
+---+----------+------+---------+--------+-----+------------+
| 1  | Per A   | 1    | 0       | 0      | 1   | 1          |
| 2  | Per B   | 1    | 1       | 2      | 0   | 2          |
| 3  | Per C   | 1    | 0       | 0      | 1   | 1          |
+----+---------+------+---------+--------+-----+------------+

如何获得此输出？

我尝试了这种方法，但我有更多这样的列要计算，它们属于子表。所以我得到的查询结果很慢

select Pid,Name,Surname,
  (select count(*) from Childs where Persons.Pid=Childs.Pid) ChildCount, 
  (select count(*) from Childs where Persons.Pid=Childs.Pid and School='Primary') Primary 
from Persons

---

您可以使用

join

和条件聚合来执行此操作：

select p.Pid, p.Name,
       sum(case when c.school = 'High' then 1 else 0 end) as high,
       sum(case when c.school = 'Primary' then 1 else 0 end) as primary,
       sum(case when c.sex = 'Man' then 1 else 0 end) as Man,
       sum(case when c.sex = 'Woman' then 1 else 0 end) as Woman,
       count(*) as ChildCount
from persons p left join
     childs c
     on p.pid = c.pid
group by p.Pid, p.Name;

试试这个：

select Pid,Name,Surname,
ifNull((select count(*) from Childs where Persons.Pid=Childs.Pid),0) ChildCount,
ifNull((select count(*) from Childs where Persons.Pid=Childs.Pid AND School='High' GROUP By Childs.Pid),0) High,
ifNull((select count(*) from Childs where Persons.Pid=Childs.Pid AND School='Primary' GROUP By Childs.Pid),0) 'primary',
ifNull((select count(*) from Childs where Persons.Pid=Childs.Pid AND Sex='Woman' GROUP By Childs.Pid),0) Woman,
ifNull((select count(*) from Childs where Persons.Pid=Childs.Pid AND Sex='Man' GROUP By Childs.Pid),0) Man

from Persons;

---

加入表格。执行分组依据。我只想使用Pid进行分组，但SQL需要其他列，如名称、姓氏等。如果在mysql中没有问题，但在mssql问题中较新的mysql版本不允许您执行无效的分组依据-除非在兼容模式下。我只想使用Pid进行分组，并想选择名称、姓氏，但SQL需要其他列，如名称，姓氏等，因为名字相同。我们怎么能做到呢