博客
关于我
强烈建议你试试无所不能的chatGPT,快点击我
Hive 学习(七) Hive之常用内置函数二
阅读量:5368 次
发布时间:2019-06-15

本文共 24992 字,大约阅读时间需要 83 分钟。

  

  

  

  

  

  

  

  

  

  

    

    

    

  

    

    

    

    

  

 

 

 

 

正文

一,引言

  不同于普通的SQL函数,Hive支持一些其他sql不支持的函数,如表生成函数和窗口函数,集合函数等等,接下来对应一一解答。

二,表生成函数

  2.1 行转列函数(explode)

  explode:可以将集合的数据遍历出来,遍历出的每一个元素为新的一行。注意,使用explode会生成一个新的表:

  如下示例:

select * from t_stu_subject;+-------------------+---------------------+-----------------------------+--+| t_stu_subject.id  | t_stu_subject.name  |   t_stu_subject.subjects    |+-------------------+---------------------+-----------------------------+--+| 1                 | zhangsan            | ["化学","物理","数学","语文"]       || 2                 | lisi                | ["化学","数学","生物","生理","卫生"]  || 3                 | wangwu              | ["化学","语文","英语","体育","生物"]  |+-------------------+---------------------+-----------------------------+--+3 rows selected (0.176 seconds)-->对subjects字段进行行转列select explode(subjects) from t_stu_subject;+------+--+| col  |+------+--+| 化学   || 物理   || 数学   || 语文   || 化学   || 数学   || 生物   || 生理   || 卫生   || 化学   || 语文   || 英语   || 体育   || 生物   |+------+--+-->错误语句:explode生成的是一个表,所以下面会报错select id,name,explode(subjects) from t_stu_subject;Error: Error while compiling statement: FAILED: SemanticException [Error 10081]: UDTF's are not supported outside the SELECT clause, nor nested in expressions (state=42000,code=10081)

  2.2 表生成函数(lateral view explode)

  lateral view用于和split、explode等UDTF一起使用的,能将一行数据拆分成多行数据,在此基础上可以对拆分的数据进行聚合,lateral view首先为原始表的每行调用UDTF,UDTF会把一行拆分成一行或者多行,lateral view在把结果组合,产生一个支持别名表的虚拟表。

  上面的explode使用看起来毫无意义,但是我们可以结合lateral view来一起使用,如下:

select id,name,sub from t_stu_subject lateral view explode(subjects) tmp as sub;+-----+-----------+------+--+| id  |   name    | sub  |+-----+-----------+------+--+| 1   | zhangsan  | 化学   || 1   | zhangsan  | 物理   || 1   | zhangsan  | 数学   || 1   | zhangsan  | 语文   || 2   | lisi      | 化学   || 2   | lisi      | 数学   || 2   | lisi      | 生物   || 2   | lisi      | 生理   || 2   | lisi      | 卫生   || 3   | wangwu    | 化学   || 3   | wangwu    | 语文   || 3   | wangwu    | 英语   || 3   | wangwu    | 体育   || 3   | wangwu    | 生物   |+-----+-----------+------+--+

  SQL代码解析:

理解: lateral view 相当于两个表在join左表:是原表右表:是explode(某个集合字段)之后产生的表而且:这个join只在同一行的数据间进行

  实例二:求word count

实例一: ==== 利用explode和lateral view 实现hive版的wordcount 有以下数据:a b c d e f ga b ce f g ab c d b对数据建表:create table t_juzi(line string) row format delimited;导入数据:load data local inpath '/root/words.txt' into table t_juzi;** ***** ******** ***** ******** ***** ******** wordcount查询语句:***** ******** ***** ******** ***** ******** select a.word,count(1) cntfrom (select tmp.* from t_juzi lateral view explode(split(line,' ')) tmp as word) agroup by a.wordorder by cnt desc;+---------+------+--+| a.word  | cnt  |+---------+------+--+| b       | 4    || c       | 3    || a       | 3    || g       | 2    || f       | 2    || e       | 2    || d       | 2    |+---------+------+--+

三,集合函数

  3.1 判断值是否存在某集合(array_contains)

  array_contains:语法结构

array_contains(Array
, value) 返回boolean值

  示例:

-->源数据查看select * from t_stu_subject;+-------------------+---------------------+-----------------------------+--+| t_stu_subject.id  | t_stu_subject.name  |   t_stu_subject.subjects    |+-------------------+---------------------+-----------------------------+--+| 1                 | zhangsan            | ["化学","物理","数学","语文"]       || 2                 | lisi                | ["化学","数学","生物","生理","卫生"]  || 3                 | wangwu              | ["化学","语文","英语","体育","生物"]  |+-------------------+---------------------+-----------------------------+--+3 rows selected (0.066 seconds)-->array_contains使用select id, name, array_contains(subjects, '语文') from t_stu_subject;+-----+-----------+--------+--+| id  |   name    |  _c2   |+-----+-----------+--------+--+| 1   | zhangsan  | true   || 2   | lisi      | false  || 3   | wangwu    | true   |+-----+-----------+--------+--+3 rows selected (13.573 seconds)

  3.2 集合排序(sort_array)

  sort_array:语法结构

sort_array(Array
) 返回排序后的数组

  示例:

select sort_array(array(3,2,6));+----------+--+|   _c0    |+----------+--+| [2,3,6]  |+----------+--+1 row selected (12.599 seconds)

  3.3 集合长度(size)

-->数据查询select * from t_stu_subject;+-------------------+---------------------+-----------------------------+--+| t_stu_subject.id  | t_stu_subject.name  |   t_stu_subject.subjects    |+-------------------+---------------------+-----------------------------+--+| 1                 | zhangsan            | ["化学","物理","数学","语文"]       || 2                 | lisi                | ["化学","数学","生物","生理","卫生"]  || 3                 | wangwu              | ["化学","语文","英语","体育","生物"]  |+-------------------+---------------------+-----------------------------+--+3 rows selected (0.069 seconds)-->size测试select id, name, size(subjects) as sub_num from t_stu_subject;+-----+-----------+----------+--+| id  |   name    | sub_num  |+-----+-----------+----------+--+| 1   | zhangsan  | 4        || 2   | lisi      | 5        || 3   | wangwu    | 5        |+-----+-----------+----------+--+3 rows selected (13.578 seconds)

  3.4 Map集合的keys值返回

  语法格式:

map_keys(Map
)

  实例:

select * from t_family;+--------------+----------------+----------------------------------------------------------------+---------------+--+| t_family.id  | t_family.name  |                    t_family.family_members                     | t_family.age  |+--------------+----------------+----------------------------------------------------------------+---------------+--+| 1            | zhangsan       | {"father":"xiaoming","mother":"xiaohuang","brother":"xiaoxu"}  | 28            || 2            | lisi           | {"father":"mayun","mother":"huangyi","brother":"guanyu"}       | 22            || 3            | wangwu         | {"father":"wangjianlin","mother":"ruhua","sister":"jingtian"}  | 29            || 4            | mayun          | {"father":"mayongzhen","mother":"angelababy"}                  | 26            |+--------------+----------------+----------------------------------------------------------------+---------------+--+-- 查出每个人有哪些亲属关系select id,name,map_keys(family_members) as relations,agefrom  t_family;+-----+-----------+--------------------------------+------+--+| id  |   name    |           relations            | age  |+-----+-----------+--------------------------------+------+--+| 1   | zhangsan  | ["father","mother","brother"]  | 28   || 2   | lisi      | ["father","mother","brother"]  | 22   || 3   | wangwu    | ["father","mother","sister"]   | 29   || 4   | mayun     | ["father","mother"]            | 26   |+-----+-----------+--------------------------------+------+--+4 rows selected (0.129 seconds)

  3.5 Map集合的values值返回

  语法结构:map_values(Map<T,T>)

  实例:

-- 查出每个人的亲人名字select id,name,map_values(family_members) as relations,agefrom  t_family;+-----+-----------+-------------------------------------+------+--+| id  |   name    |              relations              | age  |+-----+-----------+-------------------------------------+------+--+| 1   | zhangsan  | ["xiaoming","xiaohuang","xiaoxu"]   | 28   || 2   | lisi      | ["mayun","huangyi","guanyu"]        | 22   || 3   | wangwu    | ["wangjianlin","ruhua","jingtian"]  | 29   || 4   | mayun     | ["mayongzhen","angelababy"]         | 26   |+-----+-----------+-------------------------------------+------+--+4 rows selected (0.132 seconds)

四,Json解析函数

  4.1 get_json_object

  作用:解析json字符串对象,通过 '$.key' 来获取json串的value值

  语法格式:

get_json_object(json字符串,'$.key')

  实例:

select get_json_object('{"key1":3333, "key2": 4444}', '$.key1');+-------+--+|  _c0  |+-------+--+| 3333  |+-------+--+

  4.2 json_tuple

  作用:将json字符串的value值进行提取

  语法格式:

json_tuple(json字符串,key值1,key值2) as (key1, key2)

  实例:

select json_tuple('{"key1":3333, "key2": 4444}', 'key1', 'key2') as (key1, key2);+-------+-------+--+| key1  | key2  |+-------+-------+--+| 3333  | 4444  |+-------+-------+--+

五,窗口分析函数

  5.1 聚会函数+over

  hive中的窗口函数和sql中的窗口函数相类似,都是用来做一些数据分析类的工作,一般用于olap分析(在线分析处理)。

  我们都知道在sql中有一类函数叫做聚合函数,例如sum()、avg()、max()等等,这类函数可以将多行数据按照规则聚集为一行,一般来讲聚集后的行数是要少于聚集前的行数的.但是有时我们想要既显示聚集前的数据,又要显示聚集后的数据,这时我们便引入了窗口函数。

  在深入研究over字句之前,一定要注意:在SQL处理中,窗口函数都是最后一步执行,而且仅位于Order by字句之前。

  数据准备:

-->表创建create table t_order(name string, orderdate string, cost int)row format delimitedfields terminated by ',';-->数据导入load data local inpath '/root/hiveData/t_order.txt' into table t_order;-->数据展示1: jdbc:hive2://localhost:10000> select * from t_order;+---------------+--------------------+---------------+--+| t_order.name  | t_order.orderdate  | t_order.cost  |+---------------+--------------------+---------------+--+| jack          | 2015-01-01         | 10            || tony          | 2015-01-02         | 15            || jack          | 2015-02-03         | 23            || tony          | 2015-01-04         | 29            || jack          | 2015-01-05         | 46            || jack          | 2015-04-06         | 42            || tony          | 2015-01-07         | 50            || jack          | 2015-01-08         | 55            || mart          | 2015-04-08         | 62            || mart          | 2015-04-09         | 68            || neil          | 2015-05-10         | 12            || mart          | 2015-04-11         | 75            || neil          | 2015-06-12         | 80            || mart          | 2015-04-13         | 94            |+---------------+--------------------+---------------+--+14 rows selected (0.159 seconds)

  假如说我们想要查询在2015年4月份购买过的顾客及总人数,我们便可以使用窗口函数去去实现

  实例:

1: jdbc:hive2://localhost:10000> select name,count(*) over() from t_order where substring(orderdate, 1, 7)='2015-04';+-------+-----------------+--+| name  | count_window_0  |+-------+-----------------+--+| mart  | 5               || mart  | 5               || mart  | 5               || mart  | 5               || jack  | 5               |+-------+-----------------+--+5 rows selected (1.857 seconds) -->注意:正常情况下,使用count函数是必须结合分组使用,但这里配合over可以显示聚合后的数据

  可见其实在2015年4月一共有5次购买记录,mart购买了4次,jack购买了1次.事实上,大多数情况下,我们是只看去重后的结果的.针对于这种情况,我们有两种实现方式:

--->distinct方式1: jdbc:hive2://localhost:10000> select distinct name,count(*) over ()1: jdbc:hive2://localhost:10000> from t_order1: jdbc:hive2://localhost:10000> where substring(orderdate,1,7) = '2015-04';--->group by方式1: jdbc:hive2://localhost:10000> select name,count(*) over ()1: jdbc:hive2://localhost:10000> from t_order1: jdbc:hive2://localhost:10000> where substring(orderdate,1,7) = '2015-04'1: jdbc:hive2://localhost:10000> group by name;-->输出结果都为:这里就体现出了over是在后面执行的+-------+-----------------+--+| name  | count_window_0  |+-------+-----------------+--+| mart  | 2               || jack  | 2               |+-------+-----------------+--+2 rows selected (2.889 seconds)

  5.1.1 partition by 字句

  over子句之后第一个提到的就是partition by。partition by子句也可以称为查询分区子句,非常类似于Group By,都是将数据按照边界值分组,而Over之前的函数在每一个分组之内进行,如果超出了分组,则函数会重新计算.

  需求:我们想要去看顾客的购买明细及月购买总额。

1: jdbc:hive2://localhost:10000> select name,orderdate,cost,sum(cost) over(partition by month(orderdate))  # 这里month(orderdate) 提取出月份1: jdbc:hive2://localhost:10000> from t_order;+-------+-------------+-------+---------------+--+| name  |  orderdate  | cost  | sum_window_0  |+-------+-------------+-------+---------------+--+| jack  | 2015-01-01  | 10    | 205           || jack  | 2015-01-08  | 55    | 205           || tony  | 2015-01-07  | 50    | 205           || jack  | 2015-01-05  | 46    | 205           || tony  | 2015-01-04  | 29    | 205           || tony  | 2015-01-02  | 15    | 205           || jack  | 2015-02-03  | 23    | 23            || mart  | 2015-04-13  | 94    | 341           || jack  | 2015-04-06  | 42    | 341           || mart  | 2015-04-11  | 75    | 341           || mart  | 2015-04-09  | 68    | 341           || mart  | 2015-04-08  | 62    | 341           || neil  | 2015-05-10  | 12    | 12            || neil  | 2015-06-12  | 80    | 80            |+-------+-------------+-------+---------------+--+14 rows selected (1.56 seconds)

  可以看出数据已经按照月进行汇总了.

  5.1.2 order by字句

  order by子句会让输入的数据强制排序(文章前面提到过,窗口函数是SQL语句最后执行的函数,因此可以把SQL结果集想象成输入数据)。Order By子句对于诸如Row_Number(),Lead(),LAG()等函数是必须的,因为如果数据无序,这些函数的结果就没有任何意义。因此如果有了Order By子句,则Count(),Min()等计算出来的结果就没有任何意义。

  实例:

-->假如我们想要将cost按照月进行累加.这时我们引入order by子句.1: jdbc:hive2://localhost:10000> select name,orderdate,cost,sum(cost) over(partition by month(orderdate) order by orderdate )1: jdbc:hive2://localhost:10000> from t_order;+-------+-------------+-------+---------------+--+| name  |  orderdate  | cost  | sum_window_0  |+-------+-------------+-------+---------------+--+| jack  | 2015-01-01  | 10    | 10            || tony  | 2015-01-02  | 15    | 25            || tony  | 2015-01-04  | 29    | 54            || jack  | 2015-01-05  | 46    | 100           || tony  | 2015-01-07  | 50    | 150           || jack  | 2015-01-08  | 55    | 205           || jack  | 2015-02-03  | 23    | 23            || jack  | 2015-04-06  | 42    | 42            || mart  | 2015-04-08  | 62    | 104           || mart  | 2015-04-09  | 68    | 172           || mart  | 2015-04-11  | 75    | 247           || mart  | 2015-04-13  | 94    | 341           || neil  | 2015-05-10  | 12    | 12            || neil  | 2015-06-12  | 80    | 80            |+-------+-------------+-------+---------------+--+14 rows selected (1.7 seconds)-->从上面可以看出,对月进行分组切排序

  5.1.3 window 子句

  我们在上面已经通过使用partition by子句将数据进行了分组的处理.如果我们想要更细粒度的划分,我们就要引入window子句了

  我们首先要理解两个概念: 

  如果只使用partition by子句,未指定order by的话,我们的聚合是分组内的聚合. 
  使用了order by子句,未使用window子句的情况下,默认从起点到当前行.当同一个select查询中存在多个窗口函数时,他们相互之间是没有影响的.每个窗口函数应用自己的规则.

window子句: - PRECEDING:往前 - FOLLOWING:往后 - CURRENT ROW:当前行 - UNBOUNDED:起点,UNBOUNDED PRECEDING 表示从前面的起点, UNBOUNDED FOLLOWING:表示到后面的终点

  实例:

-->我们按照name进行分区,按照购物时间进行排序,做cost的累加. 如下我们结合使用window子句进行查询select name,orderdate,cost,sum(cost) over() as sample1,--所有行相加sum(cost) over(partition by name) as sample2,--按name分组,组内数据相加sum(cost) over(partition by name order by orderdate) as sample3,--按name分组,组内数据累加sum(cost) over(partition by name order by orderdate rows between UNBOUNDED PRECEDING and current row )  as sample4 ,--和sample3一样,由起点到当前行的聚合sum(cost) over(partition by name order by orderdate rows between 1 PRECEDING   and current row) as sample5, --当前行和前面一行做聚合sum(cost) over(partition by name order by orderdate rows between 1 PRECEDING   AND 1 FOLLOWING  ) as sample6,--当前行和前边一行及后面一行sum(cost) over(partition by name order by orderdate rows between current row and UNBOUNDED FOLLOWING ) as sample7 --当前行及后面所有行from t_order;+-------+-------------+-------+----------+----------+----------+----------+----------+----------+----------+--+| name  |  orderdate  | cost  | sample1  | sample2  | sample3  | sample4  | sample5  | sample6  | sample7  |+-------+-------------+-------+----------+----------+----------+----------+----------+----------+----------+--+| jack  | 2015-01-01  | 10    | 661      | 176      | 10       | 10       | 10       | 56       | 176      || jack  | 2015-01-05  | 46    | 661      | 176      | 56       | 56       | 56       | 111      | 166      || jack  | 2015-01-08  | 55    | 661      | 176      | 111      | 111      | 101      | 124      | 120      || jack  | 2015-02-03  | 23    | 661      | 176      | 134      | 134      | 78       | 120      | 65       || jack  | 2015-04-06  | 42    | 661      | 176      | 176      | 176      | 65       | 65       | 42       || mart  | 2015-04-08  | 62    | 661      | 299      | 62       | 62       | 62       | 130      | 299      || mart  | 2015-04-09  | 68    | 661      | 299      | 130      | 130      | 130      | 205      | 237      || mart  | 2015-04-11  | 75    | 661      | 299      | 205      | 205      | 143      | 237      | 169      || mart  | 2015-04-13  | 94    | 661      | 299      | 299      | 299      | 169      | 169      | 94       || neil  | 2015-05-10  | 12    | 661      | 92       | 12       | 12       | 12       | 92       | 92       || neil  | 2015-06-12  | 80    | 661      | 92       | 92       | 92       | 92       | 92       | 80       || tony  | 2015-01-02  | 15    | 661      | 94       | 15       | 15       | 15       | 44       | 94       || tony  | 2015-01-04  | 29    | 661      | 94       | 44       | 44       | 44       | 94       | 79       || tony  | 2015-01-07  | 50    | 661      | 94       | 94       | 94       | 79       | 79       | 50       |+-------+-------------+-------+----------+----------+----------+----------+----------+----------+----------+--+14 rows selected (4.959 seconds)

  5.2 分析函数

  5.2.1 ntile

  功能:用于将分组数据按顺序切分成n片,返回当前切片值

  注意:  

ntile不支持 rows between

  使用实例:

-->假如我们想要每位顾客购买金额前1/3的交易记录,我们便可以使用这个函数.select name,orderdate,cost,    ntile(3) over() as sample1 , --全局数据切片    ntile(3) over(partition by name), -- 按照name进行分组,在分组内将数据切成3份    ntile(3) over(order by cost),--全局按照cost升序排列,数据切成3份    ntile(3) over(partition by name order by cost ) --按照name分组,在分组内按照cost升序排列,数据切成3份from t_order;+-------+-------------+-------+----------+------+------+-----------------+--+| name  |  orderdate  | cost  | sample1  | _c4  | _c5  | ntile_window_3  |+-------+-------------+-------+----------+------+------+-----------------+--+| jack  | 2015-01-01  | 10    | 3        | 1    | 1    | 1               || jack  | 2015-02-03  | 23    | 3        | 1    | 1    | 1               || jack  | 2015-04-06  | 42    | 2        | 2    | 2    | 2               || jack  | 2015-01-05  | 46    | 2        | 2    | 2    | 2               || jack  | 2015-01-08  | 55    | 2        | 3    | 2    | 3               || mart  | 2015-04-08  | 62    | 2        | 1    | 2    | 1               || mart  | 2015-04-09  | 68    | 1        | 2    | 3    | 1               || mart  | 2015-04-11  | 75    | 1        | 3    | 3    | 2               || mart  | 2015-04-13  | 94    | 1        | 1    | 3    | 3               || neil  | 2015-05-10  | 12    | 1        | 2    | 1    | 1               || neil  | 2015-06-12  | 80    | 1        | 1    | 3    | 2               || tony  | 2015-01-02  | 15    | 3        | 2    | 1    | 1               || tony  | 2015-01-04  | 29    | 3        | 3    | 1    | 2               || tony  | 2015-01-07  | 50    | 2        | 1    | 2    | 3               |+-------+-------------+-------+----------+------+------+-----------------+--+14 rows selected (5.981 seconds)

  如上述数据,我们去sample4 = 1的那部分数据就是我们要的结果

  5.2.2 row_number,rank,dense_rank

  这三个窗口函数的使用场景非常多 

  row_number()从1开始,按照顺序,生成分组内记录的序列,row_number()的值不会存在重复,当排序的值相同时,按照表中记录的顺序进行排列 
  rank() 生成数据项在分组中的排名,排名相等会在名次中留下空位 
  dense_rank() 生成数据项在分组中的排名,排名相等会在名次中不会留下空位

  注意: 

  rank和dense_rank的区别在于排名相等时会不会留下空位.

  实例如下:

  5.2.3 lag和lead

  这两个函数为常用的窗口函数,可以返回上下数据行的数据. 

  实例:

-->以我们的订单表为例,假如我们想要查看顾客上次的购买时间可以这样去查询select name,orderdate,cost,lag(orderdate,1,'1900-01-01') over(partition by name order by orderdate ) as time1,lag(orderdate,2) over (partition by name order by orderdate) as time2from t_order;+-------+-------------+-------+-------------+-------------+--+| name  |  orderdate  | cost  |    time1    |    time2    |+-------+-------------+-------+-------------+-------------+--+| jack  | 2015-01-01  | 10    | 1900-01-01  | NULL        || jack  | 2015-01-05  | 46    | 2015-01-01  | NULL        || jack  | 2015-01-08  | 55    | 2015-01-05  | 2015-01-01  || jack  | 2015-02-03  | 23    | 2015-01-08  | 2015-01-05  || jack  | 2015-04-06  | 42    | 2015-02-03  | 2015-01-08  || mart  | 2015-04-08  | 62    | 1900-01-01  | NULL        || mart  | 2015-04-09  | 68    | 2015-04-08  | NULL        || mart  | 2015-04-11  | 75    | 2015-04-09  | 2015-04-08  || mart  | 2015-04-13  | 94    | 2015-04-11  | 2015-04-09  || neil  | 2015-05-10  | 12    | 1900-01-01  | NULL        || neil  | 2015-06-12  | 80    | 2015-05-10  | NULL        || tony  | 2015-01-02  | 15    | 1900-01-01  | NULL        || tony  | 2015-01-04  | 29    | 2015-01-02  | NULL        || tony  | 2015-01-07  | 50    | 2015-01-04  | 2015-01-02  |+-------+-------------+-------+-------------+-------------+--+14 rows selected (1.6 seconds)

  5.2.4 first_value和last_value

  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,orderdate,cost,first_value(orderdate) over(partition by name order by orderdate) as time1,last_value(orderdate) over(partition by name order by orderdate) as time2from t_order;+-------+-------------+-------+-------------+-------------+--+| name  |  orderdate  | cost  |    time1    |    time2    |+-------+-------------+-------+-------------+-------------+--+| jack  | 2015-01-01  | 10    | 2015-01-01  | 2015-01-01  || jack  | 2015-01-05  | 46    | 2015-01-01  | 2015-01-05  || jack  | 2015-01-08  | 55    | 2015-01-01  | 2015-01-08  || jack  | 2015-02-03  | 23    | 2015-01-01  | 2015-02-03  || jack  | 2015-04-06  | 42    | 2015-01-01  | 2015-04-06  || mart  | 2015-04-08  | 62    | 2015-04-08  | 2015-04-08  || mart  | 2015-04-09  | 68    | 2015-04-08  | 2015-04-09  || mart  | 2015-04-11  | 75    | 2015-04-08  | 2015-04-11  || mart  | 2015-04-13  | 94    | 2015-04-08  | 2015-04-13  || neil  | 2015-05-10  | 12    | 2015-05-10  | 2015-05-10  || neil  | 2015-06-12  | 80    | 2015-05-10  | 2015-06-12  || tony  | 2015-01-02  | 15    | 2015-01-02  | 2015-01-02  || tony  | 2015-01-04  | 29    | 2015-01-02  | 2015-01-04  || tony  | 2015-01-07  | 50    | 2015-01-02  | 2015-01-07  |+-------+-------------+-------+-------------+-------------+--+14 rows selected (1.588 seconds)

  5.3 扩展

  row_number的用途非常广泛,排序最好用它,它会为查询出来的每一行记录生成一个序号,依次排序且不会重复,注意使用row_number函数时必须要用over子句选择对某一列进行排序才能生成序号。

  rank函数用于返回结果集的分区内每行的排名,行的排名是相关行之前的排名数加一。简单来说rank函数就是对查询出来的记录进行排名,与row_number函数不同的是,rank函数考虑到了over子句中排序字段值相同的情况,如果使用rank函数来生成序号,over子句中排序字段值相同的序号是一样的,后面字段值不相同的序号将跳过相同的排名号排下一个,也就是相关行之前的排名数加一,可以理解为根据当前的记录数生成序号,后面的记录依此类推。

  dense_rank函数的功能与rank函数类似,dense_rank函数在生成序号时是连续的,而rank函数生成的序号有可能不连续。dense_rank函数出现相同排名时,将不跳过相同排名号,rank值紧接上一次的rank值。在各个分组内,rank()是跳跃排序,有两个第一名时接下来就是第四名,dense_rank()是连续排序,有两个第一名时仍然跟着第二名。

  借助实例能更直观地理解:

  假设现在有一张学生表student,学生表中有姓名、分数、课程编号

1: jdbc:hive2://localhost:10000> select * from student;+-------------+---------------+----------------+-----------------+--+| student.id  | student.name  | student.score  | student.course  |+-------------+---------------+----------------+-----------------+--+| 5           | elic          | 70             | 1               || 4           | dock          | 100            | 1               || 3           | clark         | 80             | 1               || 2           | bob           | 90             | 1               || 1           | alce          | 60             | 1               || 10          | jacky         | 80             | 2               || 9           | iris          | 60             | 2               || 8           | hill          | 70             | 2               || 7           | grace         | 50             | 2               || 6           | frank         | 70             | 2               |+-------------+---------------+----------------+-----------------+--+10 rows selected (0.115 seconds)

  现在需要按照课程对学生的成绩进行排序:

--row_number() 顺序排序select name,course,row_number() over(partition by course order by score desc) rank from student;+--------+---------+-------+--+|  name  | course  | rank  |+--------+---------+-------+--+| dock   | 1       | 1     || bob    | 1       | 2     || clark  | 1       | 3     || elic   | 1       | 4     || alce   | 1       | 5     || jacky  | 2       | 1     || frank  | 2       | 2     || hill   | 2       | 3     || iris   | 2       | 4     || grace  | 2       | 5     |+--------+---------+-------+--+
--rank() 跳跃排序,如果有两个第一级别时,接下来是第三级别select name,course,rank() over(partition by course order by score desc) rank from student;+--------+---------+-------+--+|  name  | course  | rank  |+--------+---------+-------+--+| dock   | 1       | 1     || bob    | 1       | 2     || clark  | 1       | 3     || elic   | 1       | 4     || alce   | 1       | 5     || jacky  | 2       | 1     || frank  | 2       | 2     || hill   | 2       | 2     || iris   | 2       | 4     || grace  | 2       | 5     |+--------+---------+-------+--+
--dense_rank() 连续排序,如果有两个第一级别时,接下来是第二级别select name,course,dense_rank() over(partition by course order by score desc) rank from student;+--------+---------+-------+--+|  name  | course  | rank  |+--------+---------+-------+--+| dock   | 1       | 1     || bob    | 1       | 2     || clark  | 1       | 3     || elic   | 1       | 4     || alce   | 1       | 5     || jacky  | 2       | 1     || frank  | 2       | 2     || hill   | 2       | 2     || iris   | 2       | 3     || grace  | 2       | 4     |+--------+---------+-------+--+10 rows selected (1.635 seconds)

关于Parttion by:

  Parttion by关键字是Oracle中分析性函数的一部分,用于给结果集进行分区。它和聚合函数Group by不同的地方在于它只是将原始数据进行名次排列,能够返回一个分组中的多条记录(记录数不变),而Group by是对原始数据进行聚合统计,一般只有一条反映统计值的结果(每组返回一条)。

  TIPS:

  使用rank over()的时候,空值是最大的,如果排序字段为null, 可能造成null字段排在最前面,影响排序结果。

  可以这样: rank over(partition by course order by score desc nulls last)

总结:

  在使用排名函数的时候需要注意以下三点:

  1、排名函数必须有 OVER 子句。

  2、排名函数必须有包含 ORDER BY 的 OVER 子句。

  3、分组内从1开始排序。

 

转载于:https://www.cnblogs.com/tashanzhishi/p/10904144.html

你可能感兴趣的文章
策略模式 C#
查看>>
[模板]树状数组
查看>>
[HDU 6447][2018CCPC网络选拔赛 1010][YJJ's Salesman][离散化+线段树+DP]
查看>>
设计模式学习的好方法
查看>>
感谢Leslie Ma
查看>>
几种排序方法
查看>>
查看数据库各表的信息
查看>>
第一阶段测试题
查看>>
第二轮冲刺第五天
查看>>
图片压缩
查看>>
Hadoop-2.6.5安装
查看>>
ES6思维导图
查看>>
第四周作业
查看>>
20151121
查看>>
线段重叠 (思维好题)
查看>>
Codeforces Round #413 C. Fountains (线段树的创建、查询、更新)
查看>>
SBuild 0.1.5 发布,基于 Scala 的构建系统
查看>>
WordPress 3.5 RC3 发布
查看>>
DOM扩展札记
查看>>
primitive assembly
查看>>