在日常的数据分析工作中，我们经常需要对数据进行分组排序、计算移动平均值、统计累计求和等操作。在MySQL 8.0之前，这类需求通常需要编写复杂的子查询或连接查询才能实现。而MySQL 8.0引入的窗口函数（Window Functions）极大地简化了这类操作，让数据分析变得更加简单高效。
本文将通过通俗易懂的方式，带你全面了解MySQL 8.0中的窗口函数，包括聚合类、排名类和跨行类窗口函数的使用方法。

什么是窗口函数？

窗口函数是一种特殊的SQL函数，它能够在不改变原有行数的情况下，对查询结果的某个"窗口"（一组相关的行）进行计算，并为每一行返回一个值。

可以把窗口函数想象成：你有一张全班学生的成绩表，窗口函数允许你同时看到每个学生的成绩、他在班级中的排名、他与前一名同学的分数差等信息，而不需要改变原始数据表的行数。这个排名是新加的一行。

窗口函数的基本语法

SELECT 列1, 列2,窗口函数() OVER ([PARTITION BY 分区字段] [ORDER BY 排序字段][frame_clause]) AS 别名
FROM 表名;

PARTITION BY：将数据分成多个分区（类似于GROUP BY，但不合并行）
ORDER BY：确定分区内数据的排序方式
frame_clause：定义窗口框架，即计算范围

一、聚合类窗口函数

聚合类窗口函数可以在保留所有行的同时，计算分组的聚合值。

1. SUM() OVER()

计算分区内的总和：

-- 计算每个部门的工资总额，同时显示每个员工的详细信息
SELECT employee_id,name,department,salary,SUM(salary) OVER (PARTITION BY department) AS dept_total_salary
FROM employees;

2. AVG() OVER()

计算分区内的平均值：

-- 计算每个部门的平均工资，同时显示每个员工的详细信息
SELECT employee_id,name,department,salary,AVG(salary) OVER (PARTITION BY department) AS dept_avg_salary
FROM employees;

3. COUNT() OVER()

计算分区内的行数：

-- 计算每个部门的员工数量，同时显示每个员工的详细信息
SELECT employee_id,name,department,salary,COUNT(*) OVER (PARTITION BY department) AS dept_employee_count
FROM employees;

二、排名类窗口函数

排名类窗口函数用于为分区内的行分配排名、序号等。

1. ROW_NUMBER()

为分区内的每一行分配一个唯一的序号：1,2,3,4,5,6,7

-- 为每个部门的员工按工资从高到低编号
SELECT employee_id,name,department,salary,ROW_NUMBER() OVER (PARTITION BY department ORDER BY salary DESC) as rank_in_dept
FROM employees;

2. RANK()

计算分区内的排名，相同值会有相同排名，并跳过后续排名：1,1,3,3,5,6,7

-- 计算每个部门内的工资排名（允许并列）
SELECT employee_id,name,department,salary,RANK() OVER (PARTITION BY department ORDER BY salary DESC) as salary_rank
FROM employees;

3. DENSE_RANK()

计算分区内的排名，相同值有相同排名，但不跳过后续排名：1,1,2,2,3,4,5,6,6

-- 计算每个部门内的工资密集排名（允许并列但不跳号）
SELECT employee_id,name,department,salary,DENSE_RANK() OVER (PARTITION BY department ORDER BY salary DESC) as dense_rank
FROM employees;

4. NTILE()

将分区内的数据分成指定数量的组：

-- 将每个部门的员工按工资高低分成4个组
SELECT employee_id,name,department,salary,NTILE(4) OVER (PARTITION BY department ORDER BY salary DESC) as quartile
FROM employees;

三、跨行类窗口函数

跨行类窗口函数可以访问分区内其他行的数据。

1. LAG()

访问分区中当前行之前的数据：

-- 查看每位员工和上一名员工的工资差异
SELECT employee_id,name,department,salary,LAG(salary) OVER (PARTITION BY department ORDER BY salary DESC) as prev_salary,salary - LAG(salary) OVER (PARTITION BY department ORDER BY salary DESC) as diff_from_prev
FROM employees;

2. LEAD()

访问分区中当前行之后的数据：

-- 查看每位员工和下一名员工的工资差异
SELECT employee_id,name,department,salary,LEAD(salary) OVER (PARTITION BY department ORDER BY salary DESC) as next_salary,salary - LEAD(salary) OVER (PARTITION BY department ORDER BY salary DESC) as diff_from_next
FROM employees;

3. FIRST_VALUE()

获取分区内第一行的值：

-- 查看每位员工与部门最高工资的差异
SELECT employee_id,name,department,salary,FIRST_VALUE(salary) OVER (PARTITION BY department ORDER BY salary DESC) as highest_salary,salary - FIRST_VALUE(salary) OVER (PARTITION BY department ORDER BY salary DESC) as diff_from_highest
FROM employees;

4. LAST_VALUE()

获取分区内最后一行的值：

-- 查看每位员工与部门最低工资的差异
SELECT employee_id,name,department,salary,LAST_VALUE(salary) OVER (PARTITION BY department ORDER BY salary DESCRANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) as lowest_salary,salary - LAST_VALUE(salary) OVER (PARTITION BY department ORDER BY salary DESCRANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) as diff_from_lowest
FROM employees;

注意：使用LAST_VALUE()时需要特别注意窗口框架的定义，否则可能得不到预期结果。

窗口框架（Window Frame）详解

窗口框架定义了窗口函数计算时使用的行范围，常用的有两种：

ROWS模式：基于物理行偏移
RANGE模式：基于逻辑值偏移

示例：计算移动平均值

-- 计算每个员工与前2行、当前行、后2行共5行的平均工资
SELECT employee_id,name,department,salary,AVG(salary) OVER (PARTITION BY department ORDER BY salary DESCROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) as moving_avg
FROM employees;

实际应用场景

场景1：计算同比/环比增长率

-- 计算每月销售额与上月相比的增长率
SELECT year,month,sales,LAG(sales) OVER (ORDER BY year, month) as prev_month_sales,(sales - LAG(sales) OVER (ORDER BY year, month)) / LAG(sales) OVER (ORDER BY year, month) * 100 as growth_rate
FROM monthly_sales;

场景2：获取Top N记录

-- 获取每个部门工资前三名的员工
WITH ranked_employees AS (SELECT employee_id,name,department,salary,DENSE_RANK() OVER (PARTITION BY department ORDER BY salary DESC) as rank_in_deptFROM employees
)
SELECT * FROM ranked_employees WHERE rank_in_dept <= 3;

场景3：计算累计百分比

-- 计算每个部门工资的累计百分比
SELECT employee_id,name,department,salary,SUM(salary) OVER (PARTITION BY department ORDER BY salary DESC) as running_total,SUM(salary) OVER (PARTITION BY department) as dept_total,SUM(salary) OVER (PARTITION BY department ORDER BY salary DESC) / SUM(salary) OVER (PARTITION BY department) * 100 as cumulative_percent
FROM employees;