Join: Inner/Outer left/right?

dragoljubdragoljub MemberPosts:241Maven
edited October 2019 inHelp
I am slightly confused by the join operator.

There are 4 types. For clarity what does each join type preform?

When only adding attributes to a small number of samples to a large group should 'outer' or 'left/right' be used?

???

Thanks,
-Gagi
Tagged:

Answers

  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, MemberPosts:2,531Unicorn
    Hi,
    I think these types are equivalent to what is used in the database domain. Please take a look athttp://en.wikipedia.org/wiki/Join_(SQL).

    Greetings,
    Sebastian
  • dragoljubdragoljub MemberPosts:241Maven
    Thanks this makes sense.

    Inner Join = Only add attributes for matching IDs (intersection of IDs) (may remove some samples)

    Left/Right Join = Only adds attributes to samples contained in the left/right data sets (keeps all samples from left or right) (can add missing values)

    Outer Join = Joins all records from both datasets and adds missing values for attributes not shared between samples.

    Hope this helps someone. Maybe we should add a wiki link in the help of this operator.

    -Gagi
Sign InorRegisterto comment.