Make r2_score compatible with Scikit-Learn

python/cuml/cuml/ensemble/randomforestregressor.pyx

@@ -698,7 +698,7 @@ class RandomForestRegressor(BaseRandomForestModel,
 
         # shortcut for default accuracy metric of r^2
         if self.accuracy_metric == "r2":
-            stats = r2_score(y_m, preds, handle=self.handle)
+            stats = r2_score(y_m, preds)
             self.handle.sync()
             del y_m
             del preds_m

python/cuml/cuml/internals/mixins.py

@@ -1,5 +1,5 @@
 #
-# Copyright (c) 2021-2023, NVIDIA CORPORATION.
+# Copyright (c) 2021-2025, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -200,7 +200,7 @@ class RegressorMixin:
     )
     @api_base_return_any_skipall
     @enable_device_interop
-    def score(self, X, y, **kwargs):
+    def score(self, X, y, sample_weight=None, **kwargs):
         """
         Scoring function for regression estimators
 
@@ -209,13 +209,8 @@ def score(self, X, y, **kwargs):
         """
         from cuml.metrics.regression import r2_score
 
-        if hasattr(self, "handle"):
-            handle = self.handle
-        else:
-            handle = None
-
         preds = self.predict(X, **kwargs)
-        return r2_score(y, preds, handle=handle)
+        return r2_score(y, preds, sample_weight=sample_weight)
 
     @staticmethod
     def _more_static_tags():

python/cuml/cuml/metrics/CMakeLists.txt

@@ -19,7 +19,6 @@ add_module_gpu_default("accuracy.pyx" ${accuracy_algo} ${metrics_algo})
 add_module_gpu_default("hinge_loss.pyx" ${hinge_loss_algo} ${metrics_algo})
 add_module_gpu_default("kl_divergence.pyx" ${kl_divergence_algo} ${metrics_algo})
 add_module_gpu_default("pairwise_distances.pyx" ${pairwise_distances_algo} ${metrics_algo})
-add_module_gpu_default("regression.pyx" ${regression_algo} ${metrics_algo})
 add_module_gpu_default("trustworthiness.pyx" ${trustworthiness_algo} ${metrics_algo})
 
 rapids_cython_create_modules(