ENH: remove random_state from API when actually not used (#599)

Author: Valentin-Laurent

doc/v1_migration_guide.rst

@@ -130,6 +130,12 @@ The aggregation method and technique for combining predictions in ensemble metho
 - **v0.9**: Previously, the ``agg_function`` parameter had two usage: to aggregate predictions when setting ``ensemble=True`` in the ``predict`` method, and to specify the aggregation technique in ``JackknifeAfterBootstrapRegressor``.
 - **v1**: The ``agg_function`` parameter has been split into two distinct parameters: ``aggregate_predictions`` and ``aggregation_method``. ``aggregate_predictions`` is specific to ``CrossConformalRegressor``, and it specifies how predictions from multiple conformal regressors are aggregated when making point predictions. ``aggregation_method`` is specific to ``JackknifeAfterBootstrapRegressor``, and it specifies the aggregation technique for combining predictions across different bootstrap samples during conformalization.
 
+``random_state``
+~~~~~~~~~~~~~~~~~~
+
+- **v0.9**: This parameter was used to control the randomness of the data splitting.
+- **v1**: This parameter has been removed in cases where data splitting is now manual. Future evolutions may reintroduce it as a general purpose randomness control parameter.
+
 ``Other parameters``
 ~~~~~~~~~~~~~~~~~~~~
 No more parameters with incorrect ``None`` defaults.

mapie_v1/integration_tests/tests/test_regression.py

@@ -71,7 +71,6 @@
             "prefit": False,
             "test_size": 0.4,
             "fit_params": {"sample_weight": sample_weight_train},
-            "random_state": RANDOM_STATE,
         }
     },
     {
@@ -89,7 +88,6 @@
             "confidence_level": [0.5, 0.5],
             "conformity_score": "gamma",
             "prefit": False,
-            "random_state": RANDOM_STATE,
         }
     },
     {
@@ -113,7 +111,6 @@
                 random_state=RANDOM_STATE
             ),
             "allow_infinite_bounds": True,
-            "random_state": RANDOM_STATE,
         }
     },
     {
@@ -130,7 +127,6 @@
             "estimator": positive_predictor,
             "confidence_level": 0.9,
             "conformity_score": GammaConformityScore(),
-            "random_state": RANDOM_STATE,
             "test_size": 0.3,
             "minimize_interval_width": True
         }
@@ -358,7 +354,6 @@ def test_intervals_and_predictions_exact_equality_jackknife(params_jackknife):
             "prefit": False,
             "test_size": 0.4,
             "fit_params": {"sample_weight": sample_weight_train},
-            "random_state": RANDOM_STATE,
         },
     },
     {
@@ -379,7 +374,6 @@ def test_intervals_and_predictions_exact_equality_jackknife(params_jackknife):
             "test_size": 0.2,
             "fit_params": {"sample_weight": sample_weight},
             "minimize_interval_width": True,
-            "random_state": RANDOM_STATE,
         },
     },
     {
@@ -398,7 +392,6 @@ def test_intervals_and_predictions_exact_equality_jackknife(params_jackknife):
             "prefit": False,
             "test_size": 0.3,
             "allow_infinite_bounds": True,
-            "random_state": RANDOM_STATE,
         },
     },
     {
@@ -414,7 +407,6 @@ def test_intervals_and_predictions_exact_equality_jackknife(params_jackknife):
             "confidence_level": 0.9,
             "prefit": False,
             "test_size": 0.3,
-            "random_state": RANDOM_STATE,
             "symmetric_intervals": False,
         },
     },
@@ -457,7 +449,7 @@ def compare_model_predictions_and_intervals(
     prefit: bool = False,
     test_size: Optional[float] = None,
     sample_weight: Optional[ArrayLike] = None,
-    random_state: int = 42,
+    random_state: int = RANDOM_STATE,
 ) -> None:
 
     if test_size is not None:

mapie_v1/regression.py

@@ -62,10 +62,6 @@ class SplitConformalRegressor:
         Controls the verbosity level.
         Higher values increase the output details.
 
-    random_state : Optional[Union[int, np.random.RandomState]], default=None
-        A seed or random state instance to ensure reproducibility in any random
-        operations within the regressor.
-
     Notes
     -----
     This implementation currently uses a ShuffleSplit cross-validation scheme
@@ -100,7 +96,6 @@ def __init__(
         prefit: bool = False,
         n_jobs: Optional[int] = None,
         verbose: int = 0,
-        random_state: Optional[Union[int, np.random.RandomState]] = None,
     ) -> None:
         check_estimator_fit_predict(estimator)
         self._estimator = estimator
@@ -118,7 +113,6 @@ def __init__(
             n_jobs=n_jobs,
             verbose=verbose,
             conformity_score=self._conformity_score,
-            random_state=random_state,
         )
 
         self._alphas = transform_confidence_level_to_alpha_list(