MySQL基础 开窗函数
目录
mysql语法
数据准备
create table emp ( empno numeric(4) not null, ename varchar(10), job varchar(9), mgr numeric(4), hiredate datetime, sal numeric(7, 2), comm numeric(7, 2), deptno numeric(2) ); insert into emp values (7369, 'SMITH', 'CLERK', 7902, '1980-12-17', 800, null, 20); insert into emp values (7499, 'ALLEN', 'SALESMAN', 7698, '1981-02-20', 1600, 300, 30); insert into emp values (7521, 'WARD', 'SALESMAN', 7698, '1981-02-22', 1250, 500, 30); insert into emp values (7566, 'JONES', 'MANAGER', 7839, '1981-04-02', 2975, null, 20); insert into emp values (7654, 'MARTIN', 'SALESMAN', 7698, '1981-09-28', 1250, 1400, 30); insert into emp values (7698, 'BLAKE', 'MANAGER', 7839, '1981-05-01', 2850, null, 30); insert into emp values (7782, 'CLARK', 'MANAGER', 7839, '1981-06-09', 2450, null, 10); insert into emp values (7788, 'SCOTT', 'ANALYST', 7566, '1982-12-09', 3000, null, 20); insert into emp values (7839, 'KING', 'PRESIDENT', null, '1981-11-17', 5000, null, 10); insert into emp values (7844, 'TURNER', 'SALESMAN', 7698, '1981-09-08', 1500, 0, 30); insert into emp values (7876, 'ADAMS', 'CLERK', 7788, '1983-01-12', 1100, null, 20); insert into emp values (7900, 'JAMES', 'CLERK', 7698, '1981-12-03', 950, null, 30); insert into emp values (7902, 'FORD', 'ANALYST', 7566, '1981-12-03', 3000, null, 20); insert into emp values (7934, 'MILLER', 'CLERK', 7782, '1982-01-23', 1300, null, 10); 1.聚合函数（分组函数）
1.聚合统计逻辑
聚合统计： group by => 分组 xianyu,<1,a,xc,asd> lxy,<as,zxf,zxf,qwr,ags> 聚合函数 => 指标 xianyu,4 lxy,5 2.函数使用
group by =》 分组 聚合函数 =》 指标统计 sun avg max min count 需求： 统计每个部门有多少个人？ 查什么？ 维度：部门 指标：人数 select deptno, count(1) as cnt from emp group by deptno; 解释： count(1) 【1.代表 先放置一个假数，然后再查询】 【2.理解为按照第几个字段进行查数】 select select + 函数 => 可以校验函数是否存在 2.开窗函数
1.语法
窗口函数： 窗口 + 函数 窗口：函数运行时 计算数据集的范围 函数：运行时的函数 1.聚合函数 sun avg max min count 2.内置窗口函数 语法结构： 函数 over([partition by xxx,...] [order by xxx,...]) over() 是以谁进行开窗【table or 数据集】 partition by：以谁进行分组 【group by column】 order by：以谁进行排序【column】 2.聚合函数：多行数据 按照一定规则 进行聚合 为一行
sum avg max... 理论上：聚合后的行数 <= 聚合前的行数 【主要是看维度选取 group by 里面的字段】 需求： 既要显示 聚合前的数据 又要显示 聚合后的数据 ？ id name sal dt sal_all 1 zs 1000 2022-4 1000 2 ls 2000 2022-4 2000 3 wu 3000 2022-4 3000 4 zs 1000 2022-5 2000 5 ls 2000 2022-5 4000 6 wu 3000 2022-5 6000 数据：
服务器 每天的启动 次数
linux01,2022-04-15,1
linux01,2022-04-16,5
linux01,2022-04-17,7
linux01,2022-04-18,2
linux01,2022-04-19,3
linux01,2022-04-20,10
linux01,2022-04-21,4
统计累计问题： 创建表 create table window01( name varchar(50), dt varchar(20), cnt int ); 插入数据 insert into window01 values(linux01,2022-04-15,1); insert into window01 values(linux01,2022-04-16,5); insert into window01 values(linux01,2022-04-17,7); insert into window01 values(linux01,2022-04-18,2); insert into window01 values(linux01,2022-04-19,3); insert into window01 values(linux01,2022-04-20,10); insert into window01 values(linux01,2022-04-21,4); insert into window01 values(linux02,2022-04-18,20); insert into window01 values(linux02,2022-04-19,30); insert into window01 values(linux02,2022-04-20,10); insert into window01 values(linux02,2022-04-21,40); 需求： 每个服务器 每天 累积启动次数 select name, dt, cnt, sum(cnt) over(partition by name order by dt) as cut_all from window01; +---------+------------+------+---------+ | name | dt | cnt | cut_all | +---------+------------+------+---------+ | linux01 | 2022-04-15 | 1 | 1 | | linux01 | 2022-04-16 | 5 | 6 | | linux01 | 2022-04-17 | 7 | 13 | | linux01 | 2022-04-18 | 2 | 15 | | linux01 | 2022-04-19 | 3 | 18 | | linux01 | 2022-04-20 | 10 | 28 | | linux01 | 2022-04-21 | 4 | 32 | | linux02 | 2022-04-18 | 20 | 20 | | linux02 | 2022-04-19 | 30 | 50 | | linux02 | 2022-04-20 | 10 | 60 | | linux02 | 2022-04-21 | 40 | 100 | +---------+------------+------+---------+ 1 9 10 11 str 【字典序】 1 10 11 9 * 从1开始，1，2，3，4，5，6，7，8，9 * 从10开始，1，10…19，2，3，4，5，6，7，8，9 * 从20开始，1，10…19，2，20…29，3，4，5，6，7，8，9 * 以此类推，所有的10位数，都插入到与他们十位数位置上相等的个位数后面。 3.内置窗口函数
窗口大小 xxx between xxx and xxx 参数 (ROWS | RANGE) BETWEEN (UNBOUNDED | [num]) PRECEDING AND ([num] PRECEDING | CURRENT ROW | (UNBOUNDED | [num]) FOLLOWING) (ROWS | RANGE) BETWEEN CURRENT ROW AND (CURRENT ROW | (UNBOUNDED | [num]) FOLLOWING) (ROWS | RANGE) BETWEEN [num] FOLLOWING AND (UNBOUNDED | [num]) FOLLOWING select name, dt, cnt, sum(cnt) over(partition by name order by dt) as cut_all, -- 无边界 sum(cnt) over(partition by name order by dt rows between unbounded preceding and current row) as cut_all2, -- 前三行 + 当前行 sum(cnt) over(partition by name order by dt rows between 3 preceding and current row) as cut_all3, -- 前三行 + 当前行 + 下一行 sum(cnt) over(partition by name order by dt rows between 3 preceding and 1 following) as cut_all4, -- 上面无边界 + 下面无边界 sum(cnt) over(partition by name order by dt rows between unbounded preceding and UNBOUNDED FOLLOWING) as cut_all5 from window01; select name, dt, cnt, -- 常规分组排序求加和 sum(cnt) over(partition by name order by dt) as cut_all, -- 整张表对时间排序，然后加和，作用到整张表，理解为18号并列有两条数据 sum(cnt) over(order by dt) as cut_all2, -- 对整张表进行加和 sum(cnt) over() as cut_all3, -- 直接按照名字进分组 sum(cnt) over(partition by name) as cut_all4 from window01 order by dt; 1.partition by 不加 => 作用整张表 数仓顺序 ods不动 union all + group by select ifnull case when join group by grouping sets 【维度组合】 4.内置窗口函数
1.取值 串行
1.串行
LAG 【窗口内 向上 第n行的值 当前行向上取一行】 LAG(column [, N[, default]]) column => 列名 n => 取几行 default => 取不到给默认值 LEAD 【窗口内 向下 第n行的值 当前行向下取一行】 select name, dt, cnt, sum(cnt) over(partition by name order by dt) as cut_all, lead(dt,1,9999-99-99) over(partition by name order by dt) as lead_alias, lead(dt,1,9999-99-99) over(partition by name order by dt) as lag_alias from window01; 2.取值
FIRST_VALUE() ： 取分组内排序后 截止到当前行 第一个值 LAST_VALUE()：取分组内排序后 截止到当前行 最后一个值 select name, dt, cnt, first_value(cnt) over(partition by name order by dt) as f_value, last_value(cnt) over(partition by name order by dt) as l_value from window01; 2.排序
分组
ntile 需求： 把数据按照某个字段进行排序，把数据分成n份ntile(n) 如果不能平均分配 优先分配到编号小的里面 select name, dt, cnt, sum(cnt) over(partition by name order by dt) as cut_all, -- 平均分成n份，不能平均分，优先把多余的放到最小的里面 ntile(2) over(partition by name order by dt) as n2, ntile(3) over(partition by name order by dt) as n3 from window01 order by dt; 排序
rank : 从1开始，按照排序 相同会重复，名次会留下空位 生成组内的记录编号 row_number： 从1开始，按照排序 生成组内的记录编号 dense_rank：从1开始，按照排序 相同会重复，名次不会留下空位 生成组内的记录编号 select name, dt, cnt, sum(cnt) over(partition by name order by dt) as cut_all, rank() over(partition by name order by cnt desc) as rk, row_number() over(partition by name order by cnt desc) as rw, dense_rank() over(partition by name order by cnt desc) as d_rk from window01;