@Balachandran Karnati Collect_list uses ArrayList, so the data will be kept in the same order they were added, to do that, uou need to use SORT BY clause in a subquery, don't use ORDER BY, it will cause your query to execute in a non-distributed way. Find a simple example below: drop table collect;
create external table collect(
key string,
order_field int,
value double
)
row format delimited fields terminated by ','
stored as textfile
;
[root@sandbox ~]# cat collect1.txt
a,1,1.0
a,3,3.0
a,2,2.0
b,1,1.0
b,2,2.0
b,3,3.0
[root@sandbox ~]# cat collect2.txt
a,1,1.1
a,3,3.1
a,2,2.1
b,1,1.1
b,2,2.1
b,3,3.1
[root@sandbox ~]# hadoop fs -put collect* /apps/hive/warehouse/collect
drop table IF EXISTS collect_sorted;
create table collect_sorted as
select key, collect_list(value)
from
(select * from collect sort by key, order_field, value desc) x
group by key
;
... View more
