Hive的考试练习

建表：

create table hive1 
(name string, course string, score int) 
row format delimited fields terminated by ',';  

load data local inpath '/home/potter/hive1.txt' 
into table hive1; 

第一小题：

(1)
select name 
from hive1 group by name having min(score) > 80;

第二小题：

(2)
select name  
from hive1 where score < 60 group by name;

先建表：

CREATE TABLE `hive2` (  
  `id` int(11) NOT NULL AUTO_INCREMENT PRIMARY KEY,    
  `username` varchar(255) DEFAULT NULL,  
  `math` int(11) DEFAULT NULL, 
  `computer` int(11) DEFAULT NULL,   
  `english` int(11) DEFAULT NULL  
) ENGINE=InnoDB DEFAULT CHARSET=utf8;  


INSERT INTO `hive2` VALUES (1, 'huangbo', 34, 58,58);  
INSERT INTO `hive2` VALUES (2, 'xuzheng', 45,87,45);  
INSERT INTO `hive2` VALUES (3, 'wangbaoqiang', 76,34,89); 

行转列语句：

SELECT id, username , 'math' course ,math AS math FROM hive2 
UNION SELECT id ,username,'computer' course ,computer as computer FROM hive2 
UNION SELECT id ,username,'english' course ,english as english FROM hive2 
ORDER BY id,username,course;

自定义数据：

1,22,33,34,2013-12-11,2
2,23,44,55,2014-11-11,4
3,12,44,12,2015-1-11,5
1,22,33,34,2013-12-11,2
2,23,44,55,2014-11-11,4
3,12,44,12,2015-1-11,5
1,22,33,34,2013-12-11,2
2,23,44,55,2014-11-11,4
3,12,44,12,2015-1-11,5
1,22,33,34,2013-12-11,2
2,23,44,55,2014-11-11,4
3,12,44,12,2015-1-11,5
1,22,33,34,2013-12-11,2
2,23,44,55,2014-11-11,4
3,12,44,12,2015-1-11,5
1,22,33,34,2013-12-11,2
2,23,44,55,2014-11-11,4
3,12,44,12,2015-1-11,5
1,22,33,34,2013-12-11,2
2,23,44,55,2014-11-11,4
3,12,44,12,2015-1-11,5
1,22,33,34,2013-12-11,2

建表：

create table hive3 
(user_id string,city_id int, cat_id string, visits_id int,dt string, hour int) 
row format delimited fields terminated by ',';

load data local inpath '/home/potter/hive3.txt' 
into table hive3;

第一小题：

PV
select city_id,dt,count(cat_id) 
from hive3 where dt="2014-11-11" 
group by city_id,dt;

结果：
23      2014-11-11      10

UV
select city_id ,dt,count(distinct user_id) 
from hive3 where dt="2014-11-11" 
group by city_id,dt;

结果：
23      2014-11-11      1

第二小题：

select city_id,dt,count(cat_id) 
from hive3 where dt="2014-11-11" and city_id=23 
group by city_id,dt;

结果：
23      2014-11-11      10

建表：

create table hive4 
(id int,name string, pid int, price int) 
row format delimited fields terminated by ',';  

load data local inpath '/home/potter/hive4.txt' 
into table hive4;

查询语句：

select id, name, a.pid, a.price 
from hive4 a 
join (select pid, max(price) over (partition by pid) 
as price from hive4) b on a.pid = b.pid 
where a.price = b.price group by id, name, a.pid ,a.price;

建表语句：

create table hive5 
(id int,name string, age int, favor string) 
row format delimited fields terminated by ':';  

load data local inpath '/home/potter/hive5.txt' 
into table hive5;

查询语句：

create table hive5_1 as  
select a.id id, a.name name,a.age age, tv.favors    
from hive5 a   
lateral view explode(split(a.favor,",")) tv as favors;

select a.id, a.name, a.age, a.favors as favors 
from hive5_1 a 
join (select max(age) over (partition by favors) as age, favors 
from hive5_1 group by age, favors) b on a.age = b.age and a.favors = b.favors 
group by a.id, a.name, a.age, a.favors;