Hive窗口函数

其他函数: Hive—Hive函数

---

---

开窗

又称开窗函数

OVER()：指定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而变化

数据准备

字段为 name,orderdata,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

建表

----建表------
create table if not exists t_window(
    name string,
    orderdate date,
    cost int
)

导入数据

原始数据的已经上传到hdfs上

load data inpath '/order.csv' into table  t_window;

# 查看数据
select * from t_window;

聚合函数

常见聚合函数（count(),sum(),max(),min(),avg()……）

---全表数据cost的总和
select name,orderdate,cost,sum(cost) over () from t_window;

window子句

UNBOUNDED 起点
CURRENT ROW 当前行
n PRECEDING 往前 n 行数据
n FOLLOWING 往后 n 行数据
UNBOUNDED PRECEDING 表示从前面的起点
UNBOUNDED FOLLOWING 表示到后面的终点

select name,orderdate,cost,
       sum(cost) over (partition by name order by orderdate rows between unbounded preceding and unbounded following) as row1,---个人累计消费总和
       sum(cost) over(partition by name order by orderdate) as row2,--个人截止到当前时间的消费总和
       sum(cost) over(partition by name order by orderdate rows between unbounded preceding and current row ) as row3,--个人截止到当前时间的消费总和
       sum(cost) over(partition by name order by orderdate rows between 1 preceding and current row ) as row4,---当前消费额与上一次消费额的总和
       sum(cost) over(partition by name order by orderdate rows between 1 preceding and 1 following ) as row5, --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 row6,--当前行及后面所有行
       sum(cost) over(partition by name order by orderdate rows between current row and 3 following) as row7 --当前消费与后三次的消费总额
from t_window;

LAG(col,n,default_val) 往前第 n 行数据

分区内滞后当前行的参数值

select name,orderdate,cost,
       -----前一个日期
       lag(orderdate,1,'1990-01-01') over (partition by name order by orderdate) as row1
from t_window;

LEAD(col,n, default_val) 往后第 n 行数据

分区内当前行前导行的参数值

select name,orderdate,cost,
        ----后一个日期
       lead(orderdate,1,'1990-01-01') over (partition by name order by orderdate) as row2
from t_window;

ROW_NUMBER() 会根据顺序计算

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

---根据名字分区，消费金额降序排序
select name,orderdate,cost,
       row_number() over (partition by name order by cost desc ) as row1
from t_window;

RANK() 排序相同时会重复，总数不会变

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

select name,orderdate,cost,
       rank() over (partition by name order by cost desc )as row2
from t_window;

DENSE_RANK() 排序相同时会重复，总数会减少

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

select name,orderdate,cost,
       dense_rank() over (partition by name order by cost desc )as row3
from t_window;

first_value取分组内排序后，截止到当前行，第一个值

select name,orderdate,cost
     --------按名字分区，查询第一笔消费的时间与所有时间对比（有需求查询员工的入职时间，与此类似）
   ,first_value(orderdate)over (partition by name order by orderdate)as time1
from t_window;

last_value取分组内排序后，截止到当前行，最后一个值

select name,orderdate,cost
     --------按名字分区，查询当前行的最后一个时间
   ,last_value(orderdate)over (partition by name order by orderdate)as time2
from t_window;

NTILE(n) 数据切片函数

把有序窗口的行分发到指定数据的组中，各个组有编号，编号从 1 开始，对于每一行，NTILE 返回此行所属的组的编号
注意：n 必须为 int 类型

select name,orderdate,cost,
       ntile(3) over() as row1---把数据分成三份
from t_window;

select name,orderdate,cost,
       ntile(3) over(partition by name)as row2---按名字分组 切片成三份
from t_window;