feat: Added alpha parameter to RidgeRegression

src/safeds/ml/classical/regression/_elastic_net_regression.py

@@ -21,11 +21,9 @@ class ElasticNetRegression(Regressor):
     ----------
     alpha : float
         Controls the regularization of the model. The higher the value, the more regularized it becomes.
-
     lasso_ratio: float
-        Number between 0 and 1 that controls the ratio between Lasso- and Ridge regularization.
-        lasso_ratio=0 is essentially RidgeRegression
-        lasso_ratio=1 is essentially LassoRegression
+        Number between 0 and 1 that controls the ratio between Lasso and Ridge regularization. If 0, only Ridge
+        regularization is used. If 1, only Lasso regularization is used.
 
     Raises
     ------

src/safeds/ml/classical/regression/_ridge_regression.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import warnings
 from typing import TYPE_CHECKING
 
 from sklearn.linear_model import Ridge as sk_Ridge
@@ -13,12 +14,36 @@
 
 
 class RidgeRegression(Regressor):
-    """Ridge regression."""
+    """
+    Ridge regression.
 
-    def __init__(self) -> None:
+    Parameters
+    ----------
+    alpha : float
+        Controls the regularization of the model. The higher the value, the more regularized it becomes.
+
+    Raises
+    ------
+    ValueError
+        If alpha is negative.
+    """
+
+    def __init__(self, alpha: float = 1.0) -> None:
         self._wrapped_regressor: sk_Ridge | None = None
         self._feature_names: list[str] | None = None
         self._target_name: str | None = None
+        self._alpha = alpha
+        if self._alpha < 0:
+            raise ValueError("alpha must be non-negative")
+        if self._alpha == 0.0:
+            warnings.warn(
+                (
+                    "Setting alpha to zero makes this model equivalent to LinearRegression. You should use "
+                    "LinearRegression instead for better numerical stability."
+                ),
+                UserWarning,
+                stacklevel=2,
+            )
 
     def fit(self, training_set: TaggedTable) -> RidgeRegression:
         """
@@ -41,10 +66,10 @@ def fit(self, training_set: TaggedTable) -> RidgeRegression:
         LearningError
             If the training data contains invalid values or if the training failed.
         """
-        wrapped_regressor = sk_Ridge()
+        wrapped_regressor = sk_Ridge(alpha=self._alpha)
         fit(wrapped_regressor, training_set)
 
-        result = RidgeRegression()
+        result = RidgeRegression(alpha=self._alpha)
         result._wrapped_regressor = wrapped_regressor
         result._feature_names = training_set.features.column_names
         result._target_name = training_set.target.name

tests/safeds/ml/classical/regression/test_ridge_regression.py

@@ -0,0 +1,32 @@
+import pytest
+from safeds.data.tabular.containers import Table
+from safeds.ml.classical.regression import RidgeRegression
+
+
+def test_should_raise_if_alpha_is_negative() -> None:
+    with pytest.raises(ValueError, match="alpha must be non-negative"):
+        RidgeRegression(alpha=-1.0)
+
+
+def test_should_warn_if_alpha_is_zero() -> None:
+    with pytest.warns(
+        UserWarning,
+        match=(
+            "Setting alpha to zero makes this model equivalent to LinearRegression. You "
+            "should use LinearRegression instead for better numerical stability."
+        ),
+    ):
+        RidgeRegression(alpha=0.0)
+
+
+def test_should_pass_alpha_to_fitted_regressor() -> None:
+    regressor = RidgeRegression(alpha=1.0)
+    fitted_regressor = regressor.fit(Table.from_dict({"A": [1, 2, 4], "B": [1, 2, 3]}).tag_columns("B"))
+    assert regressor._alpha == fitted_regressor._alpha
+
+
+def test_should_pass_alpha_to_sklearn() -> None:
+    regressor = RidgeRegression(alpha=1.0)
+    fitted_regressor = regressor.fit(Table.from_dict({"A": [1, 2, 4], "B": [1, 2, 3]}).tag_columns("B"))
+    assert fitted_regressor._wrapped_regressor is not None
+    assert fitted_regressor._wrapped_regressor.alpha == fitted_regressor._alpha