• [转]SQL中 OVER(PARTITION BY) 取上一条,下一条等


    OVER(PARTITION BY)函数介绍

    开窗函数          
         Oracle从8.1.6开始提供分析函数,分析函数用于计算基于组的某种聚合值,它和聚合函数的不同之处是:对于每个组返回多行,而聚合函数对于每个组只返回一行。

          开窗函数指定了分析函数工作的数据窗口大小,这个数据窗口大小可能会随着行的变化而变化,举例如下:
    1:over后的写法:    
       over(order by salary) 按照salary排序进行累计,order by是个默认的开窗函数
       over(partition by deptno)按照部门分区

       over(partition by deptno order by salary)

    2:开窗的窗口范围:
    over(order by salary range between 5 preceding and 5 following):窗口范围为当前行数据幅度减5加5后的范围内的。

    举例:

    --sum(s)over(order by s range between 2 preceding and 2 following) 表示加2或2的范围内的求和

     select name,class,s, sum(s)over(order by s range between 2 preceding and 2 following) mm from t2
    adf        3        45        45  --45加2减2即43到47,但是s在这个范围内只有45
    asdf       3        55        55
    cfe        2        74        74
    3dd        3        78        158 --78在76到80范围内有78,80,求和得158
    fda        1        80        158
    gds        2        92        92
    ffd        1        95        190
    dss        1        95        190
    ddd        3        99        198

    gf         3        99        198

    over(order by salary rows between 5 preceding and 5 following):窗口范围为当前行前后各移动5行。

    举例:

    --sum(s)over(order by s rows between 2 preceding and 2 following)表示在上下两行之间的范围内
    select name,class,s, sum(s)over(order by s rows between 2 preceding and 2 following) mm from t2
    adf        3        45        174  (45+55+74=174)
    asdf       3        55        252   (45+55+74+78=252)
    cfe        2        74        332    (74+55+45+78+80=332)
    3dd        3        78        379    (78+74+55+80+92=379)
    fda        1        80        419
    gds        2        92        440
    ffd        1        95        461
    dss        1        95        480
    ddd        3        99        388
    gf         3        99        293
    over(order by salary range between unbounded preceding and unbounded following)或者
    over(order by salary rows between unbounded preceding and unbounded following):窗口不做限制

    3、与over函数结合的几个函数介绍

    row_number()over()、rank()over()和dense_rank()over()函数的使用

    下面以班级成绩表t2来说明其应用

    t2表信息如下:
    cfe        2        74
    dss        1        95
    ffd        1        95
    fda        1        80
    gds        2        92
    gf         3        99
    ddd        3        99
    adf        3        45
    asdf       3        55
    3dd        3        78

    select * from                                                                      
        (                                                                           
        select name,class,s,rank()over(partition by class order by s desc) mm from t2
        )                                                                           
        where mm=1;
    得到的结果是:
    dss        1        95        1
    ffd        1        95        1
    gds        2        92        1
    gf         3        99        1
    ddd        3        99        1 

    注意:
        1.在求第一名成绩的时候,不能用row_number(),因为如果同班有两个并列第一,row_number()只返回一个结果;
    select * from                                                                      
        (                                                                           
        select name,class,s,row_number()over(partition by class order by s desc) mm from t2
        )                                                                           
        where mm=1;
    1        95        1  --95有两名但是只显示一个
    2        92        1
    3        99        1 --99有两名但也只显示一个

        2.rank()和dense_rank()可以将所有的都查找出来:
    如上可以看到采用rank可以将并列第一名的都查找出来;
         rank()和dense_rank()区别:
         --rank()是跳跃排序,有两个第二名时接下来就是第四名;
    select name,class,s,rank()over(partition by class order by s desc) mm from t2
    dss        1        95        1
    ffd        1        95        1
    fda        1        80        3 --直接就跳到了第三
    gds        2        92        1
    cfe        2        74        2
    gf         3        99        1
    ddd        3        99        1
    3dd        3        78        3
    asdf       3        55        4
    adf        3        45        5
         --dense_rank()l是连续排序,有两个第二名时仍然跟着第三名
    select name,class,s,dense_rank()over(partition by class order by s desc) mm from t2
    dss        1        95        1
    ffd        1        95        1
    fda        1        80        2 --连续排序(仍为2)
    gds        2        92        1
    cfe        2        74        2
    gf         3        99        1
    ddd        3        99        1
    3dd        3        78        2
    asdf       3        55        3
    adf        3        45        4

    --sum()over()的使用
    select name,class,s, sum(s)over(partition by class order by s desc) mm from t2 --根据班级进行分数求和
    dss        1        95        190  --由于两个95都是第一名,所以累加时是两个第一名的相加
    ffd        1        95        190 
    fda        1        80        270  --第一名加上第二名的
    gds        2        92        92
    cfe        2        74        166
    gf         3        99        198
    ddd        3        99        198
    3dd        3        78        276
    asdf       3        55        331
    adf        3        45        376

    first_value() over()和last_value() over()的使用  


    --找出这三条电路每条电路的第一条记录类型和最后一条记录类型

    SELECT opr_id,res_type,
           first_value(res_type) over(PARTITION BY opr_id ORDER BY res_type) low,
           last_value(res_type) over(PARTITION BY opr_id ORDER BY res_type rows BETWEEN unbounded preceding AND unbounded following) high
      FROM rm_circuit_route
    WHERE opr_id IN ('000100190000000000021311','000100190000000000021355','000100190000000000021339')
     ORDER BY opr_id;

     

    注:rows BETWEEN unbounded preceding AND unbounded following 的使用

    --取last_value时不使用rows BETWEEN unbounded preceding AND unbounded following的结果

    SELECT opr_id,res_type,
           first_value(res_type) over(PARTITION BY opr_id ORDER BY res_type) low,
           last_value(res_type) over(PARTITION BY opr_id ORDER BY res_type) high
      FROM rm_circuit_route
     WHERE opr_id IN ('000100190000000000021311','000100190000000000021355','000100190000000000021339')
     ORDER BY opr_id;

    如下图可以看到,如果不使用

    rows BETWEEN unbounded preceding AND unbounded following,取出的last_value由于与res_type进行进行排列,因此取出的电路的最后一行记录的类型就不是按照电路的范围提取了,而是以res_type为范围进行提取了。

     

     

    在first_value和last_value中ignore nulls的使用

    数据如下:

     

    取出该电路的第一条记录,加上ignore nulls后,如果第一条是判断的那个字段是空的,则默认取下一条,结果如下所示:

     

     

    --lag() over()函数用法(取出前n行数据)
    lag(expresstion,,)
    with a as 
    (select 1 id,'a' name from dual
     union
     select 2 id,'b' name from dual
     union
     select 3 id,'c' name from dual
     union
     select 4 id,'d' name from dual
     union
     select 5 id,'e' name from dual

    select id,name,lag(id,1,'')over(order by name) from a;

    --lead() over()函数用法(取出后N行数据)

    lead(expresstion,,)
    with a as 
    (select 1 id,'a' name from dual
     union
     select 2 id,'b' name from dual
     union
     select 3 id,'c' name from dual
     union
     select 4 id,'d' name from dual
     union
     select 5 id,'e' name from dual

    select id,name,lead(id,1,'')over(order by name) from a;

    --ratio_to_report(a)函数用法 Ratio_to_report() 括号中就是分子,over() 括号中就是分母
    with a as (select 1 a from dual
               union all
    select 1 a from dual
               union  all
    select 1 a from dual
               union all
    select 2 a from dual
               union all 
    select 3 a from dual
               union all
    select 4 a from dual
               union all
    select 4 a from dual
               union all
    select 5 a from dual
               )
    select a, ratio_to_report(a)over(partition by a) b from a 
    order by a; 

    with a as (select 1 a from dual
               union all
    select 1 a from dual
               union  all
    select 1 a from dual
               union all
    select 2 a from dual
               union all 
    select 3 a from dual
               union all
    select 4 a from dual
               union all
    select 4 a from dual
               union all
    select 5 a from dual
               )
    select a, ratio_to_report(a)over() b from a --分母缺省就是整个占比
    order by a; 

    with a as (select 1 a from dual
               union all
    select 1 a from dual
               union  all
    select 1 a from dual
               union all
    select 2 a from dual
               union all 
    select 3 a from dual
               union all
    select 4 a from dual
               union all
    select 4 a from dual
               union all
    select 5 a from dual
               )
    select a, ratio_to_report(a)over() b from a
    group by a order by a;--分组后的占比

    percent_rank用法
    计算方法:所在组排名序号-1除以该组所有的行数-1,如下所示自己计算的pr1与通过percent_rank函数得到的值是一样的:
    SELECT a.deptno,
           a.ename,
           a.sal,
           a.r,
           b.n,
           (a.r-1)/(n-1) pr1,
           percent_rank() over(PARTITION BY a.deptno ORDER BY a.sal) pr2
      FROM (SELECT deptno,
                   ename,
                   sal,
                   rank() over(PARTITION BY deptno ORDER BY sal) r --计算出在组中的排名序号
              FROM emp
             ORDER BY deptno, sal) a,
           (SELECT deptno, COUNT(1) n FROM emp GROUP BY deptno) b --按部门计算每个部门的所有成员数
     WHERE a.deptno = b.deptno;


     

    cume_dist函数
    计算方法:所在组排名序号除以该组所有的行数,但是如果存在并列情况,则需加上并列的个数-1,
              如下所示自己计算的pr1与通过percent_rank函数得到的值是一样的:
    SELECT a.deptno,
           a.ename,
           a.sal,
           a.r,
           b.n,
           c.rn,
           (a.r + c.rn - 1) / n pr1,
           cume_dist() over(PARTITION BY a.deptno ORDER BY a.sal) pr2
      FROM (SELECT deptno,
                   ename,
                   sal,
                   rank() over(PARTITION BY deptno ORDER BY sal) r
              FROM emp
             ORDER BY deptno, sal) a,
           (SELECT deptno, COUNT(1) n FROM emp GROUP BY deptno) b,
           (SELECT deptno, r, COUNT(1) rn,sal
              FROM (SELECT deptno,sal,
                           rank() over(PARTITION BY deptno ORDER BY sal) r
                      FROM emp)
             GROUP BY deptno, r,sal
             ORDER BY deptno) c --c表就是为了得到每个部门员工工资的一样的个数
     WHERE a.deptno = b.deptno
       AND a.deptno = c.deptno(+)
       AND a.sal = c.sal;

     

    percentile_cont函数
    含义:输入一个百分比(该百分比就是按照percent_rank函数计算的值),返回该百分比位置的平均值
    如下,输入百分比为0.7,因为0.7介于0.6和0.8之间,因此返回的结果就是0.6对应的sal的1500加上0.8对应的sal的1600平均
    SELECT ename,
           sal,
           deptno,
           percentile_cont(0.7) within GROUP(ORDER BY sal) over(PARTITION BY deptno) "Percentile_Cont",
           percent_rank() over(PARTITION BY deptno ORDER BY sal) "Percent_Rank"
      FROM emp
     WHERE deptno IN (30, 60);

     

    若输入的百分比为0.6,则直接0.6对应的sal值,即1500
    SELECT ename,
           sal,
           deptno,
           percentile_cont(0.6) within GROUP(ORDER BY sal) over(PARTITION BY deptno) "Percentile_Cont",
           percent_rank() over(PARTITION BY deptno ORDER BY sal) "Percent_Rank"
      FROM emp
     WHERE deptno IN (30, 60);

     

    PERCENTILE_DISC函数
    功能描述:返回一个与输入的分布百分比值相对应的数据值,分布百分比的计算方法见函数CUME_DIST,如果没有正好对应的数据值,就取大于该分布值的下一个值。
    注意:本函数与PERCENTILE_CONT的区别在找不到对应的分布值时返回的替代值的计算方法不同

    SAMPLE:下例中0.7的分布值在部门30中没有对应的Cume_Dist值,所以就取下一个分布值0.83333333所对应的SALARY来替代

    SELECT ename,
           sal,
           deptno,
           percentile_disc(0.7) within GROUP(ORDER BY sal) over(PARTITION BY deptno) "Percentile_Disc",
           cume_dist() over(PARTITION BY deptno ORDER BY sal) "Cume_Dist"
      FROM emp
     WHERE deptno IN (30, 60);
     
    原文地址:https://www.cnblogs.com/cjm123/p/8033639.html
  • 相关阅读:
    初识面向对象
    Git 子模块
    至今为止项目中遇到的问题
    vuex , 简单入(liao)门(jie)
    Git rebase
    Git
    js设计模式工厂模式
    vue slot
    Jquery学习,一道笔试题:关于表格【最近记】
    闭包与this学习
  • 原文地址:https://www.cnblogs.com/zhqian/p/9140313.html
Copyright © 2020-2023  润新知