常用的聚合函数有 sum(求和) 、count(*) 记录数 、max(最大值)、min(最小值)

group by 关键字表示要进行分类聚合的字段，比如要按照班级分类进行学生人数统计，班级就应该放在group by 的后面。
with rollup是可选语法，表明是否对分类聚合后的结果进行汇总。having关键字表示对分类后的结果再进行条件筛选
having和where的区别在于，having是对聚合后的结果进行条件的筛选，而where是在聚合前就对记录进行过滤，如果逻辑允许，我们尽可能用where先过滤记录，这样因为结果集小，将对聚合的效率大大提高。

我们将要使用的student表

mysql> select * from student;
+----+--------+-------+---------+
| id | name   | score | classid |
+----+--------+-------+---------+
|  1 | 张三   |    99 |       3 |
|  2 | 李四   |    60 |       2 |
|  3 | 王五   |    88 |       2 |
|  4 | 赵6    |    22 |       1 |
+----+--------+-------+---------+
4 rows in set (0.00 sec)

查看student表中总人数
mysql> select count(1) from student;
+----------+
| count(1) |
+----------+
|        4 |
+----------+
1 row in set (0.00 sec)

统计各个班级的人数
mysql> select classid , count(1) from student group by classid;
+---------+----------+
| classid | count(1) |
+---------+----------+
|       1 |        1 |
|       2 |        2 |
|       3 |        1 |
+---------+----------+
3 rows in set (0.00 sec)

既要统计各个班级的人数，又要统计总人数
mysql> select classid , count(1) from student group by classid with rollup;
+---------+----------+
| classid | count(1) |
+---------+----------+
|       1 |        1 |
|       2 |        2 |
|       3 |        1 |
|    NULL |        4 |
+---------+----------+
4 rows in set (0.00 sec)

统计人数大于1的班级
mysql> select classid,count(1) from student group by classid having count(1)>1;
+---------+----------+
| classid | count(1) |
+---------+----------+
|       2 |        2 |
+---------+----------+
1 row in set (0.00 sec)

统计学生总分数，最高和最低分数
mysql> select sum(score),max(score),min(score) from student;
+------------+------------+------------+
| sum(score) | max(score) | min(score) |
+------------+------------+------------+
|        269 |         99 |         22 |
+------------+------------+------------+
1 row in set (0.00 sec)