分组

group by

group by的含义:将查询结果按照1个或多个字段进行分组，字段值相同的为一组

group by可用于单个字段分组，也可用于多个字段分组

select * from students;
+----+-----------+------+--------+--------+--------+-----------+
| id | name | age | height | gender | cls_id | is_delete |
+----+-----------+------+--------+--------+--------+-----------+
| 1 | 小明 | 18 | 180.00 | 女 | 1 | |
| 2 | 小月月 | 18 | 180.00 | 女 | 2 |  |
| 3 | 彭于晏 | 29 | 185.00 | 男 | 1 | |
| 4 | 刘德华 | 59 | 175.00 | 男 | 2 |  |
| 5 | 黄蓉 | 38 | 160.00 | 女 | 1 | |
| 6 | 凤姐 | 28 | 150.00 | 保密 | 2 |  |
| 7 | 王祖贤 | 18 | 172.00 | 女 | 1 |  |
| 8 | 周杰伦 | 36 | NULL | 男 | 1 | |
| 9 | 程坤 | 27 | 181.00 | 男 | 2 | |
| 10 | 刘亦菲 | 25 | 166.00 | 女 | 2 | |
| 11 | 金星 | 33 | 162.00 | 中性 | 3 |  |
| 12 | 静香 | 12 | 180.00 | 女 | 4 | |
| 13 | 周杰 | 34 | 176.00 | 女 | 5 | |
| 14 | 郭靖 | 12 | 170.00 | 男 | 4 | |
+----+-----------+------+--------+--------+--------+-----------+
select gender from students group by gender;
+--------+
| gender |
+--------+
| 男 |
| 女 |
| 中性 |
| 保密 |

根据gender字段来分组，gender字段的全部值有4个’男’,‘女’,‘中性’,‘保密’，所以分为了4组 当group by单独使用时，只显示出每组的第一条记录, 所以group by单独使用时的实际意义不大

group by + group_concat()

group_concat(字段名)可以作为一个输出字段来使用，

表示分组之后，根据分组结果，使用group_concat()来放置每一组的某字段的值的集合

select gender from students group by gender;
+--------+
| gender |
+--------+
| 男 |
| 女 |
| 中性 |
| 保密 |
+--------+
select gender,group_concat(name) from students group by gender;
+--------+-----------------------------------------------------------+
| gender | group_concat(name) |
+--------+-----------------------------------------------------------+
| 男 | 彭于晏,刘德华,周杰伦,程坤,郭靖 |
| 女 | 小明,小月月,黄蓉,王祖贤,刘亦菲,静香,周杰 |
| 中性 | 金星 |
| 保密 | 凤姐 |
+--------+-----------------------------------------------------------+
select gender,group_concat(id) from students group by gender;
+--------+------------------+
| gender | group_concat(id) |
+--------+------------------+
| 男 | 3,4,8,9,14 |
| 女 | 1,2,5,7,10,12,13 |
| 中性 | 11 |
| 保密 | 6 |
+--------+------------------+

group by + 集合函数

通过group_concat()的启发，我们既然可以统计出每个分组的某字段的值的集合，那么我们也可以通过集合函数来对这个值的集合做一些操作

select gender,group_concat(age) from students group by gender;
+--------+----------------------+
| gender | group_concat(age) |
+--------+----------------------+
| 男 | 29,59,36,27,12 |
| 女 | 18,18,38,18,25,12,34 |
| 中性 | 33 |
| 保密 | 28 |
+--------+----------------------+
分别统计性别为男/女的人年龄平均值
select gender,avg(age) from students group by gender;
+--------+----------+
| gender | avg(age) |
+--------+----------+
| 男 | 32.6000 |
| 女 | 23.2857 |
| 中性 | 33.0000 |
| 保密 | 28.0000 |
+--------+----------+
分别统计性别为男/女的人的个数
select gender,count(*) from students group by gender;
+--------+----------+
| gender | count(*) |
+--------+----------+
| 男 | 5 |
| 女 | 7 |
| 中性 | 1 |
| 保密 | 1 |
+--------+----------+

group by + having

having 条件表达式：用来分组查询后指定一些条件来输出查询结果

having作用和where一样，但having只能用于group by

select gender,count(*) from students group by gender having count(*)>2;
+--------+----------+
| gender | count(*) |
+--------+----------+
| 男 | 5 |
| 女 | 7 |
+--------+----------+

group by + with rollup

with rollup的作用是：在最后新增一行，来记录当前列里所有记录的总和

select gender,count(*) from students group by gender with rollup;
+--------+----------+
| gender | count(*) |
+--------+----------+
| 男 | 5 |
| 女 | 7 |
| 中性 | 1 |
| 保密 | 1 |
| NULL | 14 |
+--------+----------+
select gender,group_concat(age) from students group by gender with rollup;
+--------+-------------------------------------------+
| gender | group_concat(age) |
+--------+-------------------------------------------+
| 男 | 29,59,36,27,12 |
| 女 | 18,18,38,18,25,12,34 |
| 中性 | 33 |
| 保密 | 28 |
| NULL | 29,59,36,27,12,18,18,38,18,25,12,34,33,28 |
+--------+-------------------------------------------+