Hive复杂数据类型之array

定义格式如下：

create table tableName(
......
colName array<基本类型>
......
)

说明：下标从0开始，越界不报错，以null代替

案例准备：

测试数据

zhangsan	78,89,92,96
lisi	67,75,83,94
王五	23,12

create table if not exists arr1(
    name string,
    scores array<string>
)
row format delimited
fields terminated by '\t'
stored as textfile ;
load data local inpath '/data/arr1.txt' into table arr1;
select * from arr1;

结果

+--------+---------------+
|name    |scores         |
+--------+---------------+
|zhangsan|["78,89,92,96"]|
|lisi    |["67,75,83,94"]|
|王五      |["23,12"]      |
+--------+---------------+

select scores[0] from arr1;

结果

+-----------+
|c0         |
+-----------+
|78,89,92,96|
|67,75,83,94|
|23,12      |
+-----------+

说明上面这种建表方式是不对的,应该用下面这种,还需要指定数组的分隔符

create table if not exists arr2
(
    name   string,
    scores array<String>
)
    row format delimited
        fields terminated by '\t'
        collection items terminated by ','
;
load data local inpath '/data/arr1.txt' into table arr2;
select *
from arr2;

结果如下

+--------+---------------------+
|name    |scores               |
+--------+---------------------+
|zhangsan|["78","89","92","96"]|
|lisi    |["67","75","83","94"]|
|王五      |["23","12"]          |
+--------+---------------------+

select scores[0] from arr2;

结果

+--+
|c0|
+--+
|78|
|67|
|23|
+--+

查询语句：

select * from arr1;
select name,scores[1] from arr2 where size(scores) > 3;
+--------+--+
|name    |c1|
+--------+--+
|zhangsan|89|
|lisi    |75|
+--------+--+

--统计arr2中的每个人的总成绩
select scores[0]+scores[1]+nvl(scores[2],0)+nvl(scores[3],0) from arr2;

想要一种效果：也就是将数组类型的数据元素展开，换句话说，就是列转行

zhangsan	78,89,92,96
lisi	67,75,83,94
王五	23,12
将上述效果转成下面的效果，更方便统计每个人的总成绩。
zhangsan        78
zhangsan        89
zhangsan        92
zhangsan        96
lisi	67
lisi	75
lisi	83
lisi	94
王五	23
王五	12

展开函数的使用

简介

- explode：
	展开函数(UDTF函数中的一种),作用是：接受一个数据行，然后返回产生多个数据行  
- lateral view:虚拟表。
	会将UDTF函数生成的结果放到一个虚拟表中，然后这个虚拟表会和输入行进行join来达到数据聚合的目的

案例：

select explode(score) score from arr2;
+---+
|col|
+---+
|78 |
|89 |
|92 |
|96 |
|67 |
|75 |
|83 |
|94 |
|23 |
|12 |
+---+



select  name,cj from arr2 lateral view explode(scores) subview as cj;

+--------+--+
|name    |cj|
+--------+--+
|zhangsan|78|
|zhangsan|89|
|zhangsan|92|
|zhangsan|96|
|lisi    |67|
|lisi    |75|
|lisi    |83|
|lisi    |94|
|王五      |23|
|王五      |12|
+--------+--+


- 统计每个学生的总成绩：
select  name,sum(cj) from arr2 lateral view explode(scores) subview as cj
group by name;
+--------+---+
|name    |_c1|
+--------+---+
|lisi    |319|
|zhangsan|355|
|王五      |35 |
+--------+---+

还可以将展开的表创建一个新表

create table arr_tmp as
select  name,cj from arr2 lateral view explode(scores) subview as cj;

总结

• array类型的数据在建表时要加一句
• collection items terminated by ','
• array类型的数据可以用explode结合lateral view展开