CHORE: set-up mypy for v1 code, add it to CI, fix existing warnings

Makefile

@@ -1,19 +1,24 @@
 .PHONY: tests doc build
 
+mapie_v0_folder_name = mapie_v0_package
+
 lint:
 	flake8 . --exclude=doc
 
 type-check:
 	mypy mapie
 
+v1-type-check:
+	mypy mapie_v1 --exclude $(mapie_v0_folder_name)
+
 tests:
 	pytest -vs --doctest-modules mapie
 
 integration-tests-v1:
-	@pip install mapie --no-dependencies --target=./mapie_v1/integration_tests/mapie_v0_package >/dev/null 2>&1
-	@mv ./mapie_v1/integration_tests/mapie_v0_package/mapie ./mapie_v1/integration_tests/mapie_v0_package/mapiev0
-	@- export PYTHONPATH="${PYTHONPATH}:./mapie_v1/integration_tests/mapie_v0_package"; pytest -vs mapie_v1/integration_tests/tests  -k $(pattern)
-	@mv ./mapie_v1/integration_tests/mapie_v0_package/mapiev0 ./mapie_v1/integration_tests/mapie_v0_package/mapie
+	@pip install mapie --no-dependencies --target=./mapie_v1/integration_tests/$(mapie_v0_folder_name) >/dev/null 2>&1
+	@mv ./mapie_v1/integration_tests/$(mapie_v0_folder_name)/mapie ./mapie_v1/integration_tests/$(mapie_v0_folder_name)/mapiev0
+	@- export PYTHONPATH="${PYTHONPATH}:./mapie_v1/integration_tests/$(mapie_v0_folder_name)"; pytest -vs mapie_v1/integration_tests/tests  -k $(pattern)
+	@mv ./mapie_v1/integration_tests/$(mapie_v0_folder_name)/mapiev0 ./mapie_v1/integration_tests/$(mapie_v0_folder_name)/mapie
 
 coverage:
 	pytest -vs \

mapie_v1/_utils.py

@@ -1,5 +1,5 @@
 import warnings
-from typing import Union, List
+from typing import Union, List, Tuple, cast
 
 from numpy import array
 from mapie._typing import ArrayLike, NDArray
@@ -56,3 +56,10 @@ def make_intervals_single_if_single_alpha(
     if len(alphas) == 1:
         return intervals[:, :, 0]
     return intervals
+
+
+def cast_point_predictions_to_ndarray(
+    point_predictions: Union[NDArray, Tuple[NDArray, NDArray]]
+) -> NDArray:
+    # This will be useless when we split .predict and .predict_set in back-end
+    return cast(NDArray, point_predictions)

mapie_v1/classification.py

@@ -31,15 +31,15 @@ def fit(
         y_train: ArrayLike,
         fit_params: Optional[dict] = None,
     ) -> Self:
-        pass
+        return self
 
     def conformalize(
         self,
         X_conf: ArrayLike,
         y_conf: ArrayLike,
         predict_params: Optional[dict] = None,
     ) -> Self:
-        pass
+        return self
 
     def predict(self, X: ArrayLike) -> NDArray:
         """
@@ -49,7 +49,7 @@ def predict(self, X: ArrayLike) -> NDArray:
         scikit-learn classifiers
         Shape (n_samples,)
         """
-        pass
+        return np.ndarray(0)
 
     def predict_sets(
         self,
@@ -66,7 +66,7 @@ def predict_sets(
         Shape (n_samples, n_classes, confidence_level) if confidence_level
         is a list of floats
         """
-        pass
+        return np.ndarray(0)
 
 
 class CrossConformalClassifier:
@@ -89,26 +89,26 @@ def fit(
         y_train: ArrayLike,
         fit_params: Optional[dict] = None,
     ) -> Self:
-        pass
+        return self
 
     def conformalize(
         self,
         X_conf: ArrayLike,
         y_conf: ArrayLike,
         predict_params: Optional[dict] = None
     ) -> Self:
-        pass
+        return self
 
     def predict(self,
-                X: ArrayLike):
+                X: ArrayLike) -> NDArray:
         """
         Return
         -----
         Return ponctual prediction similar to predict method of
         scikit-learn classifiers
         Shape (n_samples,)
         """
-        pass
+        return np.ndarray(0)
 
     def predict_sets(
         self,
@@ -125,4 +125,4 @@ def predict_sets(
         Shape (n_samples, n_classes, confidence_level) if confidence_level
         is a list of floats
         """
-        pass
+        return np.ndarray(0)

mapie_v1/regression.py

@@ -16,7 +16,8 @@
 )
 from mapie_v1._utils import transform_confidence_level_to_alpha_list, \
     check_method_not_naive, check_cv_not_string, hash_X_y, \
-    check_if_X_y_different_from_fit, make_intervals_single_if_single_alpha
+    check_if_X_y_different_from_fit, make_intervals_single_if_single_alpha, \
+    cast_point_predictions_to_ndarray
 
 
 class SplitConformalRegressor:
@@ -237,7 +238,7 @@ def predict(
             Array of point predictions, with shape (n_samples,).
         """
         predictions = self._mapie_regressor.predict(X, alpha=None)
-        return predictions
+        return cast_point_predictions_to_ndarray(predictions)
 
 
 class CrossConformalRegressor:
@@ -439,6 +440,8 @@ def conformalize(
             predict_params=predict_params
         )
 
+        return self
+
     def predict_set(
         self,
         X: ArrayLike,
@@ -518,7 +521,10 @@ def predict(
             self._mapie_regressor._check_agg_function(aggregation_method)
             self._mapie_regressor.agg_function = aggregation_method
 
-        return self._mapie_regressor.predict(X, alpha=None, ensemble=ensemble)
+        predictions = self._mapie_regressor.predict(
+            X, alpha=None, ensemble=ensemble
+        )
+        return cast_point_predictions_to_ndarray(predictions)
 
 
 class JackknifeAfterBootstrapRegressor:
@@ -627,6 +633,7 @@ def fit(
         Self
             The fitted JackknifeAfterBootstrapRegressor instance.
         """
+        return self
 
     def conformalize(
         self,
@@ -659,6 +666,7 @@ def conformalize(
             The JackknifeAfterBootstrapRegressor instance with
             calibrated prediction intervals.
         """
+        return self
 
     def predict_set(
         self,
@@ -684,7 +692,7 @@ def predict_set(
             Prediction intervals of shape `(n_samples, 2)`,
             with lower and upper bounds for each sample.
         """
-        pass
+        return np.ndarray(0)
 
     def predict(
         self,
@@ -712,7 +720,7 @@ def predict(
         NDArray
             Array of point predictions, with shape `(n_samples,)`.
         """
-        pass
+        return np.ndarray(0)
 
 
 class ConformalizedQuantileRegressor:
@@ -824,6 +832,7 @@ def fit(
         Self
             The fitted ConformalizedQuantileRegressor instance.
         """
+        return self
 
     def conformalize(
         self,
@@ -855,6 +864,7 @@ def conformalize(
             The ConformalizedQuantileRegressor instance with calibrated
             prediction intervals.
         """
+        return self
 
     def predict_set(
         self,
@@ -892,7 +902,7 @@ def predict_set(
             Prediction intervals with shape `(n_samples, 2)`, with lower
             and upper bounds for each sample.
         """
-        pass
+        return np.ndarray(0)
 
     def predict(
         self,
@@ -911,7 +921,7 @@ def predict(
         NDArray
             Array of point predictions with shape `(n_samples,)`.
         """
-        pass
+        return np.ndarray(0)
 
 
 class GibbsConformalRegressor: