Simple outlier removal transformers #28874
Closed
Aukevanoost
started this conversation in
Ideas
Replies: 1 comment 1 reply
-
We don't have anything that changes the number of samples. It brings a lot of complications. But we've had plenty of discussions about it. Maybe most relevant would be: scikit-learn/enhancement_proposals#15 |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
TL;DR: Idea to add simple outlier-removal transformers
To whom it may concern,
I am relatively new to the whole ML scene and I found out that a lot of transformers are very complex and not very customizable. Therefore I was wondering if it makes sense to add some very simple outlier removal transformers as can be seen in the code snippet below: This is a rough sketch to give a feeling of how they might work
or perhaps the code could be changed to a more simpler approach, but less flexible
This way more domain-focused and fine-grained filtering can be applied to columns without having to write a lot of boilerplate.
Below is some pseudocode on how these OutlierRemovers can be implemented (warning, Python is not my native tongue)
I'd like to know if (and why) this hasn't been implemented yet, or if I am misunderstanding how the pipeline and tranformers works.
Any thoughts are welcome.
Cheers
Beta Was this translation helpful? Give feedback.
All reactions