feat: take pipeline's estimator into account to decide tabular vectorizer options in TabularPipeline#2152
Open
khaoulariad wants to merge 9 commits into
Conversation
GaelVaroquaux
requested changes
Jun 10, 2026
MarieSacksick
added
the
CFM sprint June 2026
MarieSacksick
requested changes
Jun 10, 2026
Contributor
|
Can you add the link to this issue that will be closed by this PR please? |
correcting doc Co-authored-by: Marie Sacksick <79304610+MarieSacksick@users.noreply.github.com>
Co-authored-by: Marie Sacksick <79304610+MarieSacksick@users.noreply.github.com>
MarieSacksick
changed the title
MarieSacksick
changed the title
Co-authored-by: Marie Sacksick <79304610+MarieSacksick@users.noreply.github.com>
MarieSacksick
approved these changes
Jun 10, 2026
MarieSacksick
left a comment
MarieSacksick
left a comment
Contributor
There was a problem hiding this comment.
Good for me, thanks :)!
and congratulations!!
waiting for someone else review.
jeromedockes
left a comment
jeromedockes
left a comment
Member
There was a problem hiding this comment.
thank you very much @khaoulariad ! just a few comments :)
| step; | ||
| - a scikit-learn estimator: the provided estimator is used as the final step. | ||
| - a scikit-learn pipeline : the whole pipeline is kept and usual pre-processing by the TableReport | ||
| is added on top, depending on the estimator in the last step of the pipeline. |
Member
There was a problem hiding this comment.
Suggested change
| is added on top, depending on the estimator in the last step of the pipeline. | |
| is added before, depending on the estimator in the last step of the pipeline. |
| """ # noqa: E501 | ||
| vectorizer = TableVectorizer(n_jobs=n_jobs) | ||
| cat_feat_kwargs = {"categorical_features": "from_dtype"} | ||
| if isinstance(estimator, Pipeline): |
Member
There was a problem hiding this comment.
I think this should come after the checks below -- it is unlikely someone will put the string "regressor" or a class at the end of a pipeline before passing it to tabular_pipeline
also please avoid local variable names that differ by a single character like estimator and estimator_. here we can have something like
if isinstance(estimator, Pipeline):
*user_transformers, estimator = estimator.steps
else:
user_transformers = ()
...
make_pipeline(TableVectorizer(), *user_transformers, estimator)| if not isinstance(estimator_, _TREE_ENSEMBLE_CLASSES): | ||
| steps.append(SquashingScaler(max_absolute_value=5)) | ||
| steps.append(estimator) | ||
| if isinstance(estimator, Pipeline): |
Member
There was a problem hiding this comment.
with the suggestion above we can do all the handling of pipelines in one place
|
|
||
| @pytest.mark.parametrize( | ||
| "learner_kind", ["regressor", "regression", "classifier", "classification"] | ||
| "learner_kind", |
Member
There was a problem hiding this comment.
not a big deal at all but in general please try to avoid changes that are unrelated to your Pull Request to make it easier to review and avoid cluttering the git history
| assert isinstance(p.named_steps["tablevectorizer"].low_cardinality, ToCategorical) | ||
|
|
||
|
|
||
| def test_skpipeline_learner(): |
Member
There was a problem hiding this comment.
Suggested change
| def test_skpipeline_learner(): | |
| def test_estimator_is_a_pipeline(): |
| original_learner = LogisticRegression() | ||
| sk_pipeline = Pipeline([("pca", PCA()), ("clf", original_learner)]) | ||
| tab_pipeline = tabular_pipeline(sk_pipeline) | ||
| assert len([element for _, element in tab_pipeline.steps]) == 5 |
Member
There was a problem hiding this comment.
Suggested change
| assert len([element for _, element in tab_pipeline.steps]) == 5 | |
| assert len(tab_pipeline.steps) == 5 |
| sk_pipeline = Pipeline([("pca", PCA()), ("clf", original_learner)]) | ||
| tab_pipeline = tabular_pipeline(sk_pipeline) | ||
| assert len([element for _, element in tab_pipeline.steps]) == 5 | ||
| tv, imputer, scaler, pca, learner = (element for _, element in tab_pipeline.steps) |
Member
There was a problem hiding this comment.
maybe
Suggested change
| tv, imputer, scaler, pca, learner = (element for _, element in tab_pipeline.steps) | |
| tv, imputer, scaler, pca, learner = tab_pipeline.named_steps.values() |
?
Member
There was a problem hiding this comment.
also you can group the steps you don't use like *_, pca, learner = ...
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Bug Fix Pull Request
Description
Addresses #1967
Checklist
How Has This Been Tested?
AI Disclosure