feat(ensemble): add various available splitters for GB [cd build]

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "scikit-lexicographical-trees"
-version = "0.0.2"
+version = "0.0.3"
 description = "A set of python modules for machine learning and data mining"
 readme = "README.rst"
 maintainers = [

setup.py

@@ -53,7 +53,7 @@
 
 
 # VERSION = sklearn.__version__
-VERSION = "0.0.2"
+VERSION = "0.0.3"
 
 # Custom clean command to remove build artifacts
 

sklearn/ensemble/_gb.py

@@ -394,6 +394,9 @@ def __init__(
         validation_fraction=0.1,
         n_iter_no_change=None,
         tol=1e-4,
+        splitter="best",
+        feature_groups=None,
+        threshold_gain=0.0015,
     ):
         self.n_estimators = n_estimators
         self.learning_rate = learning_rate
@@ -416,6 +419,9 @@ def __init__(
         self.validation_fraction = validation_fraction
         self.n_iter_no_change = n_iter_no_change
         self.tol = tol
+        self.splitter = splitter
+        self.feature_groups = feature_groups
+        self.threshold_gain = threshold_gain
 
     @abstractmethod
     def _encode_y(self, y=None, sample_weight=None):
@@ -470,7 +476,7 @@ def _fit_stage(
             # induce regression tree on the negative gradient
             tree = DecisionTreeRegressor(
                 criterion=self.criterion,
-                splitter="best",
+                splitter=self.splitter,
                 max_depth=self.max_depth,
                 min_samples_split=self.min_samples_split,
                 min_samples_leaf=self.min_samples_leaf,
@@ -480,6 +486,8 @@ def _fit_stage(
                 max_leaf_nodes=self.max_leaf_nodes,
                 random_state=random_state,
                 ccp_alpha=self.ccp_alpha,
+                threshold_gain=self.threshold_gain,
+                features_group=self.feature_groups,
             )
 
             if self.subsample < 1.0: