hive中map相关函数总结

目录

hive官方函数解释

hive官网函数大全地址:hive官网函数大全地址

Return Type Name Description
map map(key1, value1, key2, value2, ...) Creates a map with the given key/value pairs.
array map_values(Map<K.V>) Returns an unordered array containing the values of the input map.
array map_keys(Map<K.V>) Returns an unordered array containing the keys of the input map.
map<string,string> str_to_map(text[, delimiter1, delimiter2]) Splits text into key-value pairs using two delimiters. Delimiter1 separates text into K-V pairs, and Delimiter2 splits each K-V pair. Default delimiters are ',' for delimiter1 and ':' for delimiter2.
Tkey,Tvalue explode(MAP<Tkey,Tvalue> m) Explodes a map to multiple rows. Returns a row-set with a two columns (key,value) , one row for each key-value pair from the input map. (As of Hive 0.8.0.).

示例

1、map(key1, value1, key2, value2, ...)

sql 复制代码
SELECT map('name', '张三', 'age', 20, 'gender', '男') AS student;
---结果:
student	
{"age":"20","gender":"男","name":"张三"}

2、map_values(Map<K.V>)

sql 复制代码
SELECT map_keys(map('name', '张三', 'age', 20, 'gender', '男')) AS keys;
---结果:
keys
["name","age","gender"]

3、map_values(Map<K.V>)

sql 复制代码
SELECT map_values(map('name', '张三', 'age', 20, 'gender', '男')) AS values;
---结果:
values	
["张三","20","男"]

4、str_to_map(str, delimiter1, delimiter2)

str_to_map 函数用于将一个字符串转换为 Map 对象。具体来说,str_to_map 函数会将一个由键值对组成的字符串解析成一个 Map 对象,其中键和值之间使用指定的分隔符进行分隔。其中,str 是要转换的字符串,delimiter1 是键值对之间的分隔符,delimiter2 是键和值之间的分隔符。默认情况下,delimiter1 的值是 ',',delimiter2 的值是 ':'。

sql 复制代码
SELECT str_to_map('name:张三,age:20,gender:男', ',', ':') AS student;
---结果:
student	
{"age":"20","gender":"男","name":"张三"}

SELECT str_to_map('name=张三,age=20,gender=男', ',', '=') AS student;
---结果:
student	
{"age":"20","gender":"男","name":"张三"}

5、explode (map)

sql 复制代码
select explode(map('A',10,'B',20,'C',30));
select explode(map('A',10,'B',20,'C',30)) as (key,value);
select tf.* from (select 0) t lateral view explode(map('A',10,'B',20,'C',30)) tf;
select tf.* from (select 0) t lateral view explode(map('A',10,'B',20,'C',30)) tf as key,value;
---上述四个结果均为:
key     value
A       10	
B       20	
C       30

实战

给出一组学生数据,有名字,课程,等级,分数等字段,现在求每门课的情况,包含平均成绩,及这门课包含哪些学生及学生的等级

sql 复制代码
with stud as
( select  'zhang3' as name ,'优' as grade  ,'math' as course ,'88' as score  
  union all 
  select  'li4' as name ,'良' as grade  ,'math' as course ,'72' as score
    union all 
  select  'zhao6' as name ,'差' as grade  ,'math' as course ,'44' as score
    union all 
  select  'wang5' as name ,'优' as grade  ,'chinese' as course ,'80' as score
    union all 
  select  'zhao6' as name ,'优' as grade  ,'chinese' as course ,'55' as score
    union all 
  select  'tian7' as name ,'优' as grade  ,'chinese' as course ,'75' as score
)

--sql1
select course, collect_set(concat(name,':',grade)) as collect , avg(score) from stud group by course;
---结果:
course             collect                                             avg(score)	
math        ["li4:良","zhao6:差","zhang3:优"]                           68.0
chinese     ["wang5:优","tian7:优","zhao6:优"]                          70.0
----sql2
select course, concat_ws(',',collect_set(concat(name,':',grade))) as strings , avg(score) from stud group by course;
---结果:
course                      strings                                        avg(score)
math             li4:良,zhao6:差,zhang3:优                                  68.0
chinese          wang5:优,tian7:优,zhao6:优                                 70.0
----sql3
select course, str_to_map(concat_ws(',',collect_set(concat(name,':',grade))),',',':') as maps , avg(score) from stud group by course;
---结果:
course                               maps                              avg(score)	
math                 {"li4":"良","zhang3":"优","zhao6":"差"}             68.0
chinese              {"tian7":"优","wang5":"优","zhao6":"优"}            70.0

注意:

第一种sql,collect 字段的类型是array;第二种sql,strings字段的类型是string;第三种sql,maps字段的类型是map;

问题来了,能否在第二种的基础上,实现第一种和第三种的结果,且字段类型是string;

下面实现第二种转化为第三种,实际上就是map格式转换成json字符串;

sql 复制代码
with stud as
( select  'zhang3' as name ,'优' as grade  ,'math' as course ,'88' as score  
  union all 
  select  'li4' as name ,'良' as grade  ,'math' as course ,'72' as score
    union all 
  select  'zhao6' as name ,'差' as grade  ,'math' as course ,'44' as score
    union all 
  select  'wang5' as name ,'优' as grade  ,'chinese' as course ,'80' as score
    union all 
  select  'zhao6' as name ,'优' as grade  ,'chinese' as course ,'55' as score
    union all 
  select  'tian7' as name ,'优' as grade  ,'chinese' as course ,'75' as score
)

select 
course
,concat('{"',string2,'"}') as string3
from  
(select 
course
,regexp_replace(string1,'\\,','\\"\\,\\"') as string2
from  
(
select 
    course,
    concat_ws(',', collect_list(concat_ws('":"', k,v) ) ) as string1
from (
select course, str_to_map(concat_ws(',',collect_set(concat(name,':',grade))),',',':') as maps , avg(score) 
from stud group by course
)test_map_1
lateral view outer explode(maps) kv as k,v
group by course
) tt 
) tm 

---结果:
course                               string3                            	
math                 {"li4":"良","zhang3":"优","zhao6":"差"}           
chinese              {"tian7":"优","wang5":"优","zhao6":"优"}        
相关推荐
Lx3527 小时前
Hadoop数据处理优化:减少Shuffle阶段的性能损耗
大数据·hadoop
Lx3521 天前
Hadoop容错机制深度解析:保障作业稳定运行
大数据·hadoop
IT毕设梦工厂2 天前
大数据毕业设计选题推荐-基于大数据的客户购物订单数据分析与可视化系统-Hadoop-Spark-数据可视化-BigData
大数据·hadoop·数据分析·spark·毕业设计·源码·bigdata
大数据CLUB2 天前
基于spark的澳洲光伏发电站选址预测
大数据·hadoop·分布式·数据分析·spark·数据开发
计算机编程小央姐2 天前
跟上大数据时代步伐:食物营养数据可视化分析系统技术前沿解析
大数据·hadoop·信息可视化·spark·django·课程设计·食物
IT学长编程2 天前
计算机毕业设计 基于Hadoop的健康饮食推荐系统的设计与实现 Java 大数据毕业设计 Hadoop毕业设计选题【附源码+文档报告+安装调试】
java·大数据·hadoop·毕业设计·课程设计·推荐算法·毕业论文
Lx3522 天前
Hadoop数据一致性保障:处理分布式系统常见问题
大数据·hadoop
IT学长编程2 天前
计算机毕业设计 基于Hadoop豆瓣电影数据可视化分析设计与实现 Python 大数据毕业设计 Hadoop毕业设计选题【附源码+文档报告+安装调试
大数据·hadoop·python·django·毕业设计·毕业论文·豆瓣电影数据可视化分析
Dobby_052 天前
【Hadoop】Yarn:Hadoop 生态的资源操作系统
大数据·hadoop·分布式·yarn