Join using not exists

Solved!
Ankur5289
Level 3
Join using not exists

Hello everyone,

i have a two data sets and using the join recipe. I want the dataset to be populated as such :

keep only those records which are available in only in dataset1 (say left hand side) and not available in dataset 2.

I used the join condition of the two datasets' common column using "join on" as is different . But the operation is running long time . Yes its a huge dataset with 500k records.

Is there any fastest way to get such kind of results?

0 Kudos
1 Solution
Ankur5289
Level 3
Author

Hello @Manuel  Thanks for this . let me try . do you have any predefined formulae to be used in this prepare recipe?

View solution in original post

0 Kudos
5 Replies
Manuel
Dataiker Alumni

Why not do a left Join on (dataset1, dataset2) followed by a Prepare recipe that deletes all records with data on a column originally from dataset2?

0 Kudos
Ankur5289
Level 3
Author

Hello @Manuel  Thanks for this . let me try . do you have any predefined formulae to be used in this prepare recipe?

0 Kudos
Ankur5289
Level 3
Author

Great, this worked,

0 Kudos
Manuel
Dataiker Alumni

Great

0 Kudos
AshleyW
Dataiker

Hi @Ankur5289 ,

We've recently released a new version of Dataiku, 11.3, and you can now add additional output datasets to a Join recipe that contain all the unmatched rows from either a left, right, or inner join. This will let you create a dataset with only the rows that, for example, are in the left dataset and not in the right one.

Anti-joins/exclusion joins are on our roadmap.

Cheers,

Ashley

0 Kudos