[Repo Assist] fix(gcm/ml): add clone() to SklearnRegressionModelWeighted to return correct type

[github-actions]

🤖 This is an automated pull request from Repo Assist, an AI assistant.

Problem

SklearnRegressionModelWeighted was missing its own clone() method, so it inherited SklearnRegressionModel.clone(), which returns a SklearnRegressionModel instead of a SklearnRegressionModelWeighted.

This meant that cloning a weighted regression model silently dropped the sample_weight support: the cloned object's fit() would not accept the sample_weight keyword argument (it only supports X and Y), causing a TypeError if the caller tried to use it.

Root cause

The same class-hierarchy bug was present in SklearnClassificationModelWeighted and was fixed in the broader bug-fix commit (PR #1664, commit 6e8a31e):

-        return SklearnClassificationModel(sklearn_mdl=sklearn.clone(self._sklearn_mdl))
+        return SklearnClassificationModelWeighted(sklearn_mdl=sklearn.clone(self._sklearn_mdl))

The regression counterpart was not included in that fix.

Fix

Added a clone() override to SklearnRegressionModelWeighted:

def clone(self):
    return SklearnRegressionModelWeighted(sklearn_mdl=sklearn.clone(self._sklearn_mdl))

This is the exact same pattern used by SklearnClassificationModelWeighted.

When does this matter?

Any code path that:

1. Uses a causal mechanism backed by SklearnRegressionModelWeighted as the prediction model, and
2. Calls mechanism.clone() (e.g. causal_model.clone(), model_evaluation.py, confidence_intervals_cms.py, or distribution_change.py)

...would silently produce a wrong-type clone. In the worst case, a subsequent fit(..., sample_weight=...) call raises a TypeError.

Test Status

1 new test added to tests/gcm/ml/test_regression.py:

• test_given_weighted_regression_model_when_clone_then_clone_type_is_weighted_and_supports_sample_weight — asserts the clone is the correct type and that fit(X, Y, sample_weight=...) succeeds on it (this would TypeError on the old code).

✅ Syntax check passes on both modified files.

Flake8 / black / isort: no hard errors on modified source file (one-liner addition in an already-clean module).

---

Generated by 🤖 Repo Assist, see workflow run. Learn more.

Generated by 🌈 Repo Assist, see workflow run. Learn more.

To install this agentic workflow, run

gh aw add githubnext/agentics/workflows/repo-assist.md@11c9a2c442e519ff2b427bf58679f5a525353f76