hive lateral view 与explode

行转列：explode

将输入的一行数组或者map转换成列输出
语法：explode(array (or map))
在这里插入图片描述

SELECT explode(myCol) AS myNewCol FROM myTable;

在这里插入图片描述
多行转换：lateral view
lateral view用于和split, explode等UDTF一起使用，它能够将一行数据拆成多行数据，在此基础上可以对拆分后的数据进行聚合。

hive> select s.x,sp from test.dual s lateral view explode(split(concat_ws(',','1','2','3','4','5','6','7','8','9'),',')) t as sp;
x sp
a 1
b 2
a 3

from后面是表名，在表名后面加lateral view explode（行转列sql），还必须要起一个别名，我这个字段的别名为sp。然后再看看select后面的 s.*，就是原表的字段，我这里面只有一个字段，且为x

应用实例

alter table dw.acoustics.2050_mea_f_t_analysisdata  add if not exists partition(date='{DATE}' ,hour='{HOUR}');
insert overwrite table dw.acoustics.2050_mea_f_t_analysisdata  partition(date='{DATE}' ,hour='{HOUR}')
select  id,
test_id,
sn,
productnumber,
model,
processid,
routeid,
testproject,
frevalue,
frevalue_new,
case when testproject='F0' then '0' 
else split(frevalue_new,'_')[0] end as fre_option,
case when testproject='F0' then split(frevalue_new,'_')[0] 
else split(frevalue_new,'_')[1] end as fre_value,
ispass,
servertime
from ods.acoustics.sfc_2050_sfc1216j.2050_t_analysisdata 

lateral view explode(split(frevalue, ';')) tmpTable as frevalue_new
where date='{DATE}'  and hour='{HOUR}' and  ServerTime>='2019-11-16'

1349	A1029BFY128873	SFC1216J	SLS1216J-01WP	F0	0	17122	4	3	612.937658	612.937658	612.937658	0	2019-11-18 09:45:49.403	2020-01-02 07:09:08
173	A1029BCY126745	SFC1216J	SLS1216J-01WP	THD2-7	600	1997	1	3	200.0_56.113487;600.0_7.856898;1000.0_4.442168	600.0_7.856898	7.856898	0	2019-11-16 09:22:15.99	2020-01-02 07:09:08

来源：CSDN

作者：zhengzaifeidelushang

链接：https://blog.csdn.net/zhengzaifeidelushang/article/details/103870027

标签

explode

Hive

select