Hive行转列

需求： 

孙悟空    白羊座    A                                       
沙悟净    射手座    A                                白羊座，A    孙悟空|猪八戒
宋松松    白羊座    B           =======      白羊座，B    宋松松
猪八戒    白羊座    A                                射手座，A    沙悟净|小凤姐                       
小凤姐    射手座    A

思路：表的行转列

---

案例一：使用Hive的beeline：

知识点：concat(string1,string2)  string1和string2需要是列名

[root@master hive-1.2.2]# ./bin/beeline 
...
0: jdbc:hive2://master:10000> desc t_vehicle_log;

0: jdbc:hive2://master:10000> select concat(vehicle_speed,vehicle_plate) from t_vehicle_log;

--concat()
0: jdbc:hive2://master:10000> select concat(vehicle_speed,",",vehicle_plate) from t_vehicle_log;

 加上一个分隔符

---

案例二、concat_ws(）

0: jdbc:hive2://master:10000> select concat_ws(",",monitor_id,camera_id,vehicle_plate) from t_vehicle_log;

 

---

案例三、collect_set()    对某一列去重，并返回数组

0: jdbc:hive2://master:10000> select monitor_id from t_vehicle_log;

0: jdbc:hive2://master:10000> select collect_set(monitor_id) from t_vehicle_log;

 

建表：

create table person_info(
    name string,
    constellation string,
    blood_type string)
row format delimited fields terminated by "\t";


--插数据
load data local inpath '/usr/local/src/test4/hive/person_info.txt' into table person_info;

concat_ws()查询：

0: jdbc:hive2://master:10000> select concat_ws(",",constellation,blood_type) c_b,name from person_info;

结果：

 对上一步结果作为子查询，在查询：

select 
    t1.c_b, 
    collect_set(t1.name)
from(
    select concat_ws(",",constellation,blood_type) c_b,
    name from person_info
) t1
group by t1.c_b;

最终结果：collect_set()返回的是数组，concat_ws()接受的string或者是string数组

将数组划分开：

select 
    t1.c_b, 
    concat_ws("|",collect_set(t1.name))
from(
    select concat_ws(",",constellation,blood_type) c_b,
    name from person_info
) t1
group by t1.c_b;

结果：