Hive常用函数大全（二）（窗口函数、分析函数、增强group）

窗口函数与分析函数

应用场景： 
（1）用于分区排序 
（2）动态Group By 
（3）Top N 
（4）累计计算 
（5）层次查询

窗口函数

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 
LEAD(col,n,DEFAULT) ：用于统计窗口内往下第n行值。第一个参数为列名，第二个参数为往下第n行（可选，默认为1），第三个参数为默认值（当往下第n行为NULL时候，取默认值，如不指定，则为NULL） 
LAG(col,n,DEFAULT) ：与lead相反，用于统计窗口内往上第n行值。第一个参数为列名，第二个参数为往上第n行（可选，默认为1），第三个参数为默认值（当往上第n行为NULL时候，取默认值，如不指定，则为NULL）

OVER从句

1、使用标准的聚合函数COUNT、SUM、MIN、MAX、AVG 
2、使用PARTITION BY语句，使用一个或者多个原始数据类型的列 
3、使用PARTITION BY与ORDER BY语句，使用一个或者多个数据类型的分区或者排序列 
4、使用窗口规范，窗口规范支持以下格式：

(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

当ORDER BY后面缺少窗口从句条件，窗口规范默认是 RANGE BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW.

当ORDER BY和窗口从句都缺失, 窗口规范默认是 ROW BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING.

OVER从句支持以下函数， 但是并不支持和窗口一起使用它们。 
Ranking函数: Rank, NTile, DenseRank, CumeDist, PercentRank. 
Lead 和 Lag 函数.

分析函数

ROW_NUMBER() 从1开始，按照顺序，生成分组内记录的序列,比如，按照pv降序排列，生成分组内每天的pv名次,ROW_NUMBER()的应用场景非常多，再比如，获取分组内排序第一的记录;获取一个session中的第一条refer等。 
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 
CUME_DIST 小于等于当前值的行数/分组内总行数。比如，统计小于等于当前薪水的人数，所占总人数的比例 
PERCENT_RANK 分组内当前行的RANK值-1/分组内总行数-1 
NTILE(n) 用于将分组数据按照顺序切分成n片，返回当前切片值，如果切片不均匀，默认增加第一个切片的分布。NTILE不支持ROWS BETWEEN，比如 NTILE(2) OVER(PARTITION BY cookieid ORDER BY createtime ROWS BETWEEN 3 PRECEDING AND CURRENT ROW)。

Hive2.1.0及以后支持Distinct

在聚合函数（SUM, COUNT and AVG）中，支持distinct，但是在ORDER BY 或者 窗口限制不支持。

COUNT(DISTINCT a) OVER (PARTITION BY c)

Hive 2.2.0中在使用ORDER BY和窗口限制时支持distinct

COUNT(DISTINCT a) OVER (PARTITION BY c ORDER BY d ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING)

Hive2.1.0及以后支持在OVER从句中支持聚合函数

SELECT rank() OVER (ORDER BY sum(b))
FROM T
GROUP BY a;

 

## COUNT、SUM、MIN、MAX、AVG

## COUNT、SUM、MIN、MAX、AVG
select 
    user_id,
    user_type,
    sales,
    --默认为从起点到当前行
    sum(sales) OVER(PARTITION BY user_type ORDER BY sales asc) AS sales_1,
    --从起点到当前行，结果与sales_1不同。
    sum(sales) OVER(PARTITION BY user_type ORDER BY sales asc ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS sales_2,
    --当前行+往前3行
    sum(sales) OVER(PARTITION BY user_type ORDER BY sales asc ROWS BETWEEN 3 PRECEDING AND CURRENT ROW) AS sales_3,
    --当前行+往前3行+往后1行
    sum(sales) OVER(PARTITION BY user_type ORDER BY sales asc ROWS BETWEEN 3 PRECEDING AND 1 FOLLOWING) AS sales_4,
    --当前行+往后所有行  
    sum(sales) OVER(PARTITION BY user_type ORDER BY sales asc ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING) AS sales_5,
    --分组内所有行
    SUM(sales) OVER(PARTITION BY user_type) AS sales_6                          
from 
    order_detail
order by 
    user_type,
    sales,
    user_id;

result

注意:

结果和ORDER BY相关,默认为升序
如果不指定ROWS BETWEEN,默认为从起点到当前行;
如果不指定ORDER BY，则将分组内所有值累加;

关键是理解ROWS BETWEEN含义,也叫做WINDOW子句：
PRECEDING：往前
FOLLOWING：往后
CURRENT ROW：当前行
UNBOUNDED：无界限（起点或终点）
UNBOUNDED PRECEDING：表示从前面的起点 
UNBOUNDED FOLLOWING：表示到后面的终点
其他COUNT、AVG，MIN，MAX，和SUM用法一样。

 

## first_value与last_value

## first_value与last_value
select 
    user_id,
    user_type,
    ROW_NUMBER() OVER(PARTITION BY user_type ORDER BY sales) AS row_num,  
    first_value(user_id) over (partition by user_type order by sales desc) as max_sales_user,
    first_value(user_id) over (partition by user_type order by sales asc) as min_sales_user,
    last_value(user_id) over (partition by user_type order by sales desc) as curr_last_min_user,
    last_value(user_id) over (partition by user_type order by sales asc) as curr_last_max_user
from 
    order_detail;

result

 

 ## lead与lag

## lead与lag
select 
    user_id,device_id,
    lead(device_id) over (order by sales) as default_after_one_line,
    lag(device_id) over (order by sales) as default_before_one_line,
    lead(device_id,2) over (order by sales) as after_two_line,
    lag(device_id,2,'abc') over (order by sales) as before_two_line
from 
    order_detail;

result

 

## RANK、ROW_NUMBER、DENSE_RANK

## RANK、ROW_NUMBER、DENSE_RANK
select 
    user_id,user_type,sales,
    RANK() over (partition by user_type order by sales desc) as r,
    ROW_NUMBER() over (partition by user_type order by sales desc) as rn,
    DENSE_RANK() over (partition by user_type order by sales desc) as dr
from
    order_detail;

result 

 

## NTILE

## NTILE
select 
    user_type,sales,
    --分组内将数据分成2片
    NTILE(2) OVER(PARTITION BY user_type ORDER BY sales) AS nt2,
    --分组内将数据分成3片    
    NTILE(3) OVER(PARTITION BY user_type ORDER BY sales) AS nt3,
    --分组内将数据分成4片    
    NTILE(4) OVER(PARTITION BY user_type ORDER BY sales) AS nt4,
    --将所有数据分成4片
    NTILE(4) OVER(ORDER BY sales) AS all_nt4
from 
    order_detail
order by 
    user_type,
    sales;

result 

 

## 求取sale前20%的用户ID

# 求取sale前20%的用户ID
select
    user_id
from
(
    select 
        user_id,
        NTILE(5) OVER(ORDER BY sales desc) AS nt
    from 
        order_detail
)A
where nt=1;

 

## CUME_DIST、PERCENT_RANK 

## CUME_DIST、PERCENT_RANK 
select 
user_id,user_type,sales,
--没有partition,所有数据均为1组
CUME_DIST() OVER(ORDER BY sales) AS cd1,
--按照user_type进行分组
CUME_DIST() OVER(PARTITION BY user_type ORDER BY sales) AS cd2 
from 
order_detail;   

result

select 
user_type,sales
--分组内总行数      
SUM(1) OVER(PARTITION BY user_type) AS s, 
--RANK值  
RANK() OVER(ORDER BY sales) AS r,    
PERCENT_RANK() OVER(ORDER BY sales) AS pr,
--分组内     
PERCENT_RANK() OVER(PARTITION BY user_type ORDER BY sales) AS prg 
from 
order_detail;   

 

 

增强的聚合 Cube和Grouping 和Rollup

这几个分析函数通常用于OLAP中，不能累加，而且需要根据不同维度上钻和下钻的指标统计，比如，分小时、天、月的UV数。

## GROUPING SETS

 在一个GROUP BY查询中，根据不同的维度组合进行聚合，等价于将不同维度的GROUP BY结果集进行UNION ALL, 
其中的GROUPING__ID，表示结果属于哪一个分组集合。

select
    user_type,
    sales,
    count(user_id) as pv,
    GROUPING__ID 
from 
    order_detail
group by 
    user_type,sales
GROUPING SETS(user_type,sales) 
ORDER BY 
    GROUPING__ID;

result 

 

select
    user_type,
    sales,
    count(user_id) as pv,
    GROUPING__ID 
from 
    order_detail
group by 
    user_type,sales
GROUPING SETS(user_type,sales,(user_type,sales)) 
ORDER BY 
    GROUPING__ID;

result

 

## CUBE 

根据GROUP BY的维度的所有组合进行聚合。

select
    user_type,
    sales,
    count(user_id) as pv,
    GROUPING__ID 
from 
    order_detail
group by 
    user_type,sales
WITH CUBE 
ORDER BY 
    GROUPING__ID;

 result

 

## ROLLUP 

是CUBE的子集，以最左侧的维度为主，从该维度进行层级聚合。

select
    user_type,
    sales,
    count(user_id) as pv,
    GROUPING__ID 
from 
    order_detail
group by 
    user_type,sales
WITH ROLLUP 
ORDER BY 
    GROUPING__ID;

result

 

转载：https://blog.csdn.net/scgaliguodong123_/article/details/60135385