py-why / py-why/EconML

Question about EconML's S-Learner implementation

Open
#457 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Jupyter Notebook
Stars
4.8k
Forks
827
PR merge metrics
No merged PRs in 30d

Description

Hi, I'm looking at the SLearner Meta Learner implementation, and in the code there's a note describing that the treatment for control column is added to the feature set too. Could anyone explain why we need to do this? For example, thinking of a binary treatment value, after the OHE we just need one column (as the note itself explains, since it can be linearly derived from the other).

Was it for any abstraction/implementation reason? Asking this because in some tests of reimplementing a Slearner the model was outputting a different result than EconML implementation, but we weren't adding this control column like EconML does.

https://github.com/microsoft/EconML/blob/9630421315cd90f4a44bdd6dcba493c2c5de6bc9/econml/metalearners/_metalearners.py#L170

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reading econml/metalearners/_metalearners.py around line 170 and compare the S-Learner behavior with the reimplementation tests described in the issue. Done would require a confirmed explanation of the control column's purpose and, if appropriate, documenting that rationale or identifying the relevant implementation behavior.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Documentation
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.