• MySql 分组查询


    分组函数

    功能:用作统计使用,又称为聚合函数或统计函数或组函数

    分类:
    sum 求和、avg 平均值、max 最大值 、min 最小值 、count 计算个数

    特点:
    1、sum、avg一般用于处理数值型
    max、min、count可以处理任何类型
    2、以上分组函数都忽略null值

    3、可以和distinct搭配实现去重的运算

    4、count函数的单独介绍
    一般使用count(*)用作统计行数

    5、和分组函数一同查询的字段要求是group by后的字段

    #1、简单 的使用
    SELECT SUM(salary) FROM employees;
    SELECT AVG(salary) FROM employees;
    SELECT MIN(salary) FROM employees;
    SELECT MAX(salary) FROM employees;
    SELECT COUNT(salary) FROM employees;
    
    
    SELECT SUM(salary) 和,AVG(salary) 平均,MAX(salary) 最高,MIN(salary) 最低,COUNT(salary) 个数
    FROM employees;
    
    SELECT SUM(salary) 和,ROUND(AVG(salary),2) 平均,MAX(salary) 最高,MIN(salary) 最低,COUNT(salary) 个数
    FROM employees;
    
    #2、参数支持哪些类型
    
    SELECT SUM(last_name) ,AVG(last_name) FROM employees;
    SELECT SUM(hiredate) ,AVG(hiredate) FROM employees;
    
    SELECT MAX(last_name),MIN(last_name) FROM employees;
    
    SELECT MAX(hiredate),MIN(hiredate) FROM employees;
    
    SELECT COUNT(commission_pct) FROM employees;
    SELECT COUNT(last_name) FROM employees;
    
    #3、是否忽略null
    
    SELECT SUM(commission_pct) ,AVG(commission_pct),SUM(commission_pct)/35,SUM(commission_pct)/107 FROM employees;
    
    SELECT MAX(commission_pct) ,MIN(commission_pct) FROM employees;
    
    SELECT COUNT(commission_pct) FROM employees;
    SELECT commission_pct FROM employees;
    
    
    #4、和distinct搭配
    
    SELECT SUM(DISTINCT salary),SUM(salary) FROM employees;
    
    SELECT COUNT(DISTINCT salary),COUNT(salary) FROM employees;
    
    
    
    #5、count函数的详细介绍
    
    SELECT COUNT(salary) FROM employees;
    
    
    SELECT COUNT(*) FROM employees;
    
    SELECT COUNT(1) FROM employees;
    
    效率:
    MYISAM存储引擎下  ,COUNT(*)的效率高
    INNODB存储引擎下,COUNT(*)和COUNT(1)的效率差不多,比COUNT(字段)要高一些
    
    
    #6、和分组函数一同查询的字段有限制
    
    SELECT AVG(salary),employee_id  FROM employees;
    分组函数

    select 查询列表
    from 表
    【where 筛选条件】
    group by 分组的字段
    【order by 排序的字段】;

    特点:
    1、和分组函数一同查询的字段必须是group by后出现的字段
    2、筛选分为两类:分组前筛选和分组后筛选
    针对的表 位置 连接的关键字
    分组前筛选 原始表 group by前 where

    分组后筛选 group by后的结果集 group by后 having

    问题1:分组函数做筛选能不能放在where后面
    答:不能

    问题2:where——group by——having

    一般来讲,能用分组前筛选的,尽量使用分组前筛选,提高效率

    3、分组可以按单个字段也可以按多个字段
    4、可以搭配着排序使用

    #引入:查询每个部门的员工个数

    SELECT COUNT(*) FROM employees WHERE department_id=90;

    #1.简单的分组 

    #案例1:查询每个工种的员工平均工资

    SELECT AVG(salary),job_id
    FROM employees
    GROUP BY job_id;

    #案例2:查询每个位置的部门个数

    SELECT COUNT(*),location_id
    FROM departments
    GROUP BY location_id;

    #2、可以实现分组前的筛选

    #案例1:查询邮箱中包含a字符的 每个部门的最高工资

    SELECT MAX(salary),department_id
    FROM employees
    WHERE email LIKE '%a%'
    GROUP BY department_id;

    #案例2:查询有奖金的每个领导手下员工的平均工资

    SELECT AVG(salary),manager_id
    FROM employees
    WHERE commission_pct IS NOT NULL
    GROUP BY manager_id;

    #3、分组后筛选

    #案例:查询哪个部门的员工个数>5

    #①查询每个部门的员工个数

    SELECT COUNT(*),department_id
    FROM employees
    GROUP BY department_id;

    #② 筛选刚才①结果

    SELECT COUNT(*),department_id
    FROM employees
    GROUP BY department_id
    HAVING COUNT(*)>5;

    #案例2:每个工种有奖金的员工的最高工资>12000的工种编号和最高工资

    SELECT job_id,MAX(salary)
    FROM employees
    WHERE commission_pct IS NOT NULL
    GROUP BY job_id
    HAVING MAX(salary)>12000;

    #案例3:领导编号>102的每个领导手下的最低工资大于5000的领导编号和最低工资

    manager_id>102
    
    SELECT manager_id,MIN(salary)
    FROM employees
    GROUP BY manager_id
    HAVING MIN(salary)>5000;

    #4.添加排序

    #案例:每个工种有奖金的员工的最高工资>6000的工种编号和最高工资,按最高工资升序

    SELECT job_id,MAX(salary) m
    FROM employees
    WHERE commission_pct IS NOT NULL
    GROUP BY job_id
    HAVING m>6000
    ORDER BY m ;

    #5.按多个字段分组

    #案例:查询每个工种每个部门的最低工资,并按最低工资降序

    SELECT MIN(salary),job_id,department_id
    FROM employees
    GROUP BY department_id,job_id
    ORDER BY MIN(salary) DESC;
  • 相关阅读:
    GridSearchCV.grid_scores_和mean_validation_score报错
    scikit-learn使用fetch_mldata无法下载MNIST数据集的问题
    Python数据科学手册Seaborn马拉松可视化里时分秒转化为秒数的问题
    Jupyter导出PDF从入门到绝望(已解决)
    食谱数据库数据找不到的问题
    TensorBoard计算加速
    TensorBoard可视化
    TensorFlow高层封装:从入门到喷这本书
    <!DOCTYPE html> 详解
    不同数据库之间表数据的迁移
  • 原文地址:https://www.cnblogs.com/sunjinchao/p/14063027.html
Copyright © 2020-2023  润新知