Say I do a join like this:
A = load 'inputa' as (f1,f2,f3,f4);
B = load 'inputb' as (f1,f2,f5,f6);
J = join A by (f1,f2) left outer, B by (f1,f2);
What is the simplest way to arrive at J's field names being f1,f2,f3,f4,f5,f6
? (as opposed to A:f1, A:f2, A:f3, A:f4, B:f1, B:f2, B:f5, B:f6
)
I know I can do the following:
C = foreach J generate A:f1 as f1, A:f2 as f2, A:f3 as f3, A:f4 as f4, B:f5 as f5, B:f6 as f6;
But I run into situations where I have a lot of columns following the join, so writing that all out is impractical. Is it something I should use a UDF for? I'm a pig novice and could use some guidance.
Thanks!
Aucun commentaire:
Enregistrer un commentaire