Hello all, welcome to another article on Apache Hive. In this article, we will see how can we use COLLECT_SET and COLLECT_LIST to get a list of comma separated values for particular column while doing grouping operation.
While doing hive queries we have used group by operation very often to perform all kinds of aggregation operations like sum, count, max etc. Consider there is a table with a schema like following.
We want to calculate average salary for each employee and also get a list of departments employee has worked.  If we try to run the query without dept_id in GROUP BY, Hive will complain.
 If we include dept_id in GROUP BY clause we will get a result like this.

This is because each employee may have worked in more than one department. that is why we cannot include dept_id in GROUP BY.

Use case I :

This is one of a use case where we can use COLLECT_SET and COLLECT_LIST. If we want to list all the departments for an employee we can just use COLLECT_SET which will return an array of DISTINCT dept_id for that employee.
We can also use COLLECT_LIST if we do not want to remove duplicate values from particular column.

COLLECT_SET and COLLECT_LIST return array for column type used. We can perform array operations like the size and contains.




Use case II:

This is one simple use case. But we can use COLLECT_SET and COLLECT_LIST according to needs to simplify query.

Suppose we want to know an employee has been promoted or not while getting his average salary. We can know about his promotion status from title column. If an employee has more than two titles we can say that he has been promoted.

We can write that query in following way.



In this way we can use COLLECT_SET and COLLECT_LIST to simplify queries in certain conditions.

Leave a Reply

Close Menu