scikit-learn · lorentzenchr · Jul 1, 2023 · Aug 26, 2016 · Aug 28, 2016 · Sep 2, 2016
diff --git a/doc/whats_new/v1.4.rst b/doc/whats_new/v1.4.rst
@@ -59,6 +59,31 @@ TODO: update at the time of the release.
   passed to the ``fit`` method of the the estimator. :pr:`26506` by `Adrin
   Jalali`_.
 
+
+:mod:`sklearn.ensemble`
+.......................
+
+- |Feature| :class:`ensemble.RandomForestClassifier`,
+  :class:`ensemble.RandomForestRegressor`, :class:`ensemble.ExtraTreesClassifier`
+  and :class:`ensemble.ExtraTreesRegressor` now support monotonic constraints,
+  useful when features are supposed to have a positive/negative effect on the target.
+  Missing values in the train data and multi-output targets are not supported.
+  :pr:`13649` by :user:`Samuel Ronsin <samronsin>`,
+  initiated by :user:`Patrick O'Reilly <pat-oreilly>`.
+
+
+:mod:`sklearn.tree`
+...................
+
+- |Feature| :class:`tree.DecisionTreeClassifier`, :class:`tree.DecisionTreeRegressor`,
+  :class:`tree.ExtraTreeClassifier` and :class:`tree.ExtraTreeRegressor` now support
+  monotonic constraints, useful when features are supposed to have a positive/negative
+  effect on the target. Missing values in the train data and multi-output targets are
+  not supported.
+  :pr:`13649` by :user:`Samuel Ronsin <samronsin>`, initiated by
+  :user:`Patrick O'Reilly <pat-oreilly>`.
+
+
 :mod:`sklearn.decomposition`
 ............................
 

diff --git a/sklearn/ensemble/_forest.py b/sklearn/ensemble/_forest.py
@@ -1273,6 +1273,25 @@ class RandomForestClassifier(ForestClassifier):
 
         .. versionadded:: 0.22
 
+    monotonic_cst : array-like of int of shape (n_features), default=None
+        Indicates the monotonicity constraint to enforce on each feature.
+          - 1: monotonic increase
+          - 0: no constraint
+          - -1: monotonic decrease
+
+        If monotonic_cst is None, no constraints are applied.
+
+        Monotonicity constraints are not supported for:
+          - multiclass classifications (i.e. when `n_classes > 2`),
+          - multioutput classifications (i.e. when `n_outputs_ > 1`),
+          - classifications trained on data with missing values.
+
+        The constraints hold over the probability of the positive class.
+
+        Read more in the :ref:`User Guide <monotonic_cst_gbdt>`.
+
+        .. versionadded:: 1.4
+
     Attributes
     ----------
     estimator_ : :class:`~sklearn.tree.DecisionTreeClassifier`
@@ -1413,6 +1432,7 @@ def __init__(
         class_weight=None,
         ccp_alpha=0.0,
         max_samples=None,
+        monotonic_cst=None,
     ):
         super().__init__(
             estimator=DecisionTreeClassifier(),
@@ -1428,6 +1448,7 @@ def __init__(
                 "min_impurity_decrease",
                 "random_state",
                 "ccp_alpha",
+                "monotonic_cst",
             ),
             bootstrap=bootstrap,
             oob_score=oob_score,
@@ -1447,6 +1468,7 @@ def __init__(
         self.max_features = max_features
         self.max_leaf_nodes = max_leaf_nodes
         self.min_impurity_decrease = min_impurity_decrease
+        self.monotonic_cst = monotonic_cst
         self.ccp_alpha = ccp_alpha
 
 
@@ -1627,6 +1649,22 @@ class RandomForestRegressor(ForestRegressor):
 
         .. versionadded:: 0.22
 
+    monotonic_cst : array-like of int of shape (n_features), default=None
+        Indicates the monotonicity constraint to enforce on each feature.
+          - 1: monotonically increasing
+          - 0: no constraint
+          - -1: monotonically decreasing
+
+        If monotonic_cst is None, no constraints are applied.
+
+        Monotonicity constraints are not supported for:
+          - multioutput regressions (i.e. when `n_outputs_ > 1`),
+          - regressions trained on data with missing values.
+
+        Read more in the :ref:`User Guide <monotonic_cst_gbdt>`.
+
+        .. versionadded:: 1.4
+
     Attributes
     ----------
     estimator_ : :class:`~sklearn.tree.DecisionTreeRegressor`
@@ -1754,6 +1792,7 @@ def __init__(
         warm_start=False,
         ccp_alpha=0.0,
         max_samples=None,
+        monotonic_cst=None,
     ):
         super().__init__(
             estimator=DecisionTreeRegressor(),
@@ -1769,6 +1808,7 @@ def __init__(
                 "min_impurity_decrease",
                 "random_state",
                 "ccp_alpha",
+                "monotonic_cst",
             ),
             bootstrap=bootstrap,
             oob_score=oob_score,
@@ -1788,6 +1828,7 @@ def __init__(
         self.max_leaf_nodes = max_leaf_nodes
         self.min_impurity_decrease = min_impurity_decrease
         self.ccp_alpha = ccp_alpha
+        self.monotonic_cst = monotonic_cst
 
 
 class ExtraTreesClassifier(ForestClassifier):
@@ -1975,6 +2016,25 @@ class ExtraTreesClassifier(ForestClassifier):
 
         .. versionadded:: 0.22
 
+    monotonic_cst : array-like of int of shape (n_features), default=None
+        Indicates the monotonicity constraint to enforce on each feature.
+          - 1: monotonically increasing
+          - 0: no constraint
+          - -1: monotonically decreasing
+
+        If monotonic_cst is None, no constraints are applied.
+
+        Monotonicity constraints are not supported for:
+          - multiclass classifications (i.e. when `n_classes > 2`),
+          - multioutput classifications (i.e. when `n_outputs_ > 1`),
+          - classifications trained on data with missing values.
+
+        The constraints hold over the probability of the positive class.
+
+        Read more in the :ref:`User Guide <monotonic_cst_gbdt>`.
+
+        .. versionadded:: 1.4
+
     Attributes
     ----------
     estimator_ : :class:`~sklearn.tree.ExtraTreesClassifier`
@@ -2104,6 +2164,7 @@ def __init__(
         class_weight=None,
         ccp_alpha=0.0,
         max_samples=None,
+        monotonic_cst=None,
     ):
         super().__init__(
             estimator=ExtraTreeClassifier(),
@@ -2119,6 +2180,7 @@ def __init__(
                 "min_impurity_decrease",
                 "random_state",
                 "ccp_alpha",
+                "monotonic_cst",
             ),
             bootstrap=bootstrap,
             oob_score=oob_score,
@@ -2139,6 +2201,7 @@ def __init__(
         self.max_leaf_nodes = max_leaf_nodes
         self.min_impurity_decrease = min_impurity_decrease
         self.ccp_alpha = ccp_alpha
+        self.monotonic_cst = monotonic_cst
 
 
 class ExtraTreesRegressor(ForestRegressor):
@@ -2314,6 +2377,22 @@ class ExtraTreesRegressor(ForestRegressor):
 
         .. versionadded:: 0.22
 
+    monotonic_cst : array-like of int of shape (n_features), default=None
+        Indicates the monotonicity constraint to enforce on each feature.
+          - 1: monotonically increasing
+          - 0: no constraint
+          - -1: monotonically decreasing
+
+        If monotonic_cst is None, no constraints are applied.
+
+        Monotonicity constraints are not supported for:
+          - multioutput regressions (i.e. when `n_outputs_ > 1`),
+          - regressions trained on data with missing values.
+
+        Read more in the :ref:`User Guide <monotonic_cst_gbdt>`.
+
+        .. versionadded:: 1.4
+
     Attributes
     ----------
     estimator_ : :class:`~sklearn.tree.ExtraTreeRegressor`
@@ -2426,6 +2505,7 @@ def __init__(
         warm_start=False,
         ccp_alpha=0.0,
         max_samples=None,
+        monotonic_cst=None,
     ):
         super().__init__(
             estimator=ExtraTreeRegressor(),
@@ -2441,6 +2521,7 @@ def __init__(
                 "min_impurity_decrease",
                 "random_state",
                 "ccp_alpha",
+                "monotonic_cst",
             ),
             bootstrap=bootstrap,
             oob_score=oob_score,
@@ -2460,6 +2541,7 @@ def __init__(
         self.max_leaf_nodes = max_leaf_nodes
         self.min_impurity_decrease = min_impurity_decrease
         self.ccp_alpha = ccp_alpha
+        self.monotonic_cst = monotonic_cst
 
 
 class RandomTreesEmbedding(TransformerMixin, BaseForest):
@@ -2653,7 +2735,7 @@ class RandomTreesEmbedding(TransformerMixin, BaseForest):
         **BaseDecisionTree._parameter_constraints,
         "sparse_output": ["boolean"],
     }
-    for param in ("max_features", "ccp_alpha", "splitter"):
+    for param in ("max_features", "ccp_alpha", "splitter", "monotonic_cst"):
         _parameter_constraints.pop(param)
 
     criterion = "squared_error"

diff --git a/sklearn/ensemble/_gb.py b/sklearn/ensemble/_gb.py
@@ -137,6 +137,7 @@ class BaseGradientBoosting(BaseEnsemble, metaclass=ABCMeta):
         "tol": [Interval(Real, 0.0, None, closed="left")],
     }
     _parameter_constraints.pop("splitter")
+    _parameter_constraints.pop("monotonic_cst")
 
     @abstractmethod
     def __init__(