Hive - Unpivot functionality in hive

Question

I have two table as follows:

Table A

userid | code | code_name | property_id
0001   | 1    | apple_id  | Y1234
0031   | 4    | mango_id  | G4567
0008         


        

Answer 1

You actually dont need to unpivot to get the result.

SELECT * FROM 
(
    SELECT userid,
        CASE WHEN property_id = apple_id THEN new_id  
             WHEN property_id = mango_id THEN new_id  
             WHEN property_id = grape_id THEN new_id  
             WHEN property_id = peach_id THEN new_id
        END AS newid
    FROM
    (SELECT * FROM a FULL JOIN b) X
) y
WHERE newid IS NOT NULL;

Result:

y.userid  y.newid
1         N456098
31        N002345
8         N129087
13        N109876

Answer 2

Whenever I want to pivot a table in Hive, I collect key:value pairs to a map and then reference each key in the next level, creating new columns. This is the opposite of that.

Query:

select a.userid, y.new_id
from (
  select new_id, fruit_name, fruit_code
  from (
    select new_id, map("apple_id", apple_id
                     , "mango_id", mango_id
                     , "grape_id", grape_id
                     , "peach_id", peach_id) as fruit_map
    from table_2 ) x
  lateral view explode(fruit_map) exptbl1 as fruit_name, fruit_code ) y
join table_A a
on (y.fruit_code=a.property_id)

Output:

0001    N456098
0031    N002345
0008    N129087
00013   N109876