Oracle中rank() over()函数排名

--已知：两种排名方式(分区和不分区)：使用和不使用partition

--两种计算方式(连续,不连续),对应函数：dense_rank,rank

查询原始数据：学号、姓名、科目名、成绩

select * from t_score

1、rank() over(order by 列名排序)的结果是不连续的.

如果有4个人，其中有3个是并列第1名，那么最后的排序结果结果如：1 1 1 4

select scoreid, studentid,COURSENAME,totalexamscore ,
rank() over(order by TOTALEXAMSCORE desc)orderbyNum
from SCORECOURSE a ,COURSESCORE b
where a.SCORECOURSEID = b.SCORECOURSEID

举例、查询各学生科目为Oracle排名(简单排名)

select sc.s_id,sc.s_name,sub_name,sc.score,rank() over (order by score desc) 名次from t_score scwhere sub_name='Oracle'

2、dense_rank()over(order by 列名排序)的结果是连续的.

如果有4个人，其中有3个是并列第1名，那么最后的排序结果如：1 1 1 2

select scoreid, studentid,COURSENAME,totalexamscore ,
dense_rank()over(order by TOTALEXAMSCORE desc)orderbyNum
from SCORECOURSE a ,COURSESCORE b
where a.SCORECOURSEID = b.SCORECOURSEID

对比：rank()与dense_rank()：非连续排名与连续排名(都是简单排名)

举例、dense_rank()

select sc.s_id,sc.s_name,sub_name,sc.score,dense_rank() over (order by score desc) 名次from t_score scwhere sub_name='Oracle'

3、rank () OVER (PARTITION BY 列名 ORDER BY 列名排序)

使用分区方式获取每门课程的最高分。

SELECT *   FROM (SELECT SCOREID,STUDENTID,COURSENAME,TOTALEXAMSCORE,RANK() OVER(PARTITION BY COURSENAME ORDER BY TOTALEXAMSCORE DESC) ORDERBYNUM FROM SCORECOURSE A,COURSESCORE B          WHERE A.SCORECOURSEID = B.SCORECOURSEID AND STUDENTID = 'xxxxx')  WHERE ORDERBYNUM < 2

举例 1、查询各学生各科排名(分区排名)

•分区只是将原始数据进行名次排列(记录数不变),

select sc.s_id,sc.s_name,sub_name,sc.score,rank() over(partition by sub_name order by score desc) 名次from t_score sc

举例2、查询各科前2名(分区排名)

·类似：新闻表，求栏目点击率在前3位的新闻。

商品表，求各类别销售额在前10位的商品。

select * from (select sc.s_id,sc.s_name,sub_name,sc.score,dense_rank() over (partition by sub_name order by score desc) 名次from t_score sc) xwhere x.名次<=2

4、根据总分查询各同学名次

查询各同学总分

select s_id,s_name,sum(score) sum_score from t_scoregroup by s_id,s_name

·根据总分查询各同学名次

select x.*,rank() over (order by sum_score desc) 名次from (select s_id,s_name,sum(score) sum_score from t_scoregroup by s_id,s_name ) x

-----使用over实现成绩求和
-----SUM (totalexamscore) OVER (ORDER BY studentid) sum1 实现的是连续求和，如第一个学生的总评成绩是30，则sum1就展示为30，到第二个学生成绩出现的时候，则会依次累加
-----SUM (totalexamscore) OVER () sum2 就相当于是单纯的求和，和直接使用sum是一致的

SELECT scoreid, studentid, totalexamscore,SUM (totalexamscore) OVER (ORDER BY studentid) sum1,SUM (totalexamscore) OVER () sum2FROM coursescore

注：这个案例想了很久也没想到很直观的描述，还请大家亲自去测试一下吧！！

语法：

rank() over (order by 排序字段 顺序)
rank() over (partition by 分组字段 order by 排序字段 顺序)

1.顺序：asc|desc 名次与业务相关：
示例：求优秀学员，成绩：降序迟到次数：升序
2.分区字段：根据什么字段进行分区。

问题：分区与分组有什么区别?

•分区只是将原始数据进行名次排列(记录数不变),

•分组是对原始数据进行聚合统计(记录数变少,每组返回一条)。

注意：使用rank()over(order by 排序字段顺序)排序的时候，空值是最大的
(如果排序字段为null，可能造成在排序时将null字段排在最前面，影响排序的正确性。
所以建议将dense_rank()over(order by 列名排序)改为dense_rank()over(order by 列名排序 nulls last)
这样只要排序字段为null,就会放在最后，而不会影响排序结果).

sql脚本：

create table t_score(autoid   number  primary key,s_id     number(3),s_name   char(8) not null,sub_name varchar2(20),score    number(10,2));insert into t_score (autoid, s_id, s_name, sub_name, score)values (8, 1, '张三    ', '语文', 80);insert into t_score (autoid, s_id, s_name, sub_name, score)values (9, 2, '李四    ', '数学', 80);insert into t_score (autoid, s_id, s_name, sub_name, score)values (10, 1, '张三    ', '数学', 0);insert into t_score (autoid, s_id, s_name, sub_name, score)values (11, 2, '李四    ', '语文', 50);insert into t_score (autoid, s_id, s_name, sub_name, score)values (12, 3, '张三丰  ', '语文', 10);insert into t_score (autoid, s_id, s_name, sub_name, score)values (13, 3, '张三丰  ', '数学', null);insert into t_score (autoid, s_id, s_name, sub_name, score)values (14, 3, '张三丰  ', '体育', 120);insert into t_score (autoid, s_id, s_name, sub_name, score)values (15, 4, '杨过    ', 'java', 90);insert into t_score (autoid, s_id, s_name, sub_name, score)values (16, 5, 'mike    ', 'c++', 80);insert into t_score (autoid, s_id, s_name, sub_name, score)values (3, 3, '张三丰  ', 'oracle', 0);insert into t_score (autoid, s_id, s_name, sub_name, score)values (4, 4, '杨过    ', 'oracle', 77);insert into t_score (autoid, s_id, s_name, sub_name, score)values (17, 2, '李四    ', 'oracle', 77);commit;