Add cuml.internals.validation, check_is_fitted checks

python/cuml/cuml/_thirdparty/sklearn/preprocessing/_column_transformer.py

@@ -1,6 +1,6 @@
 # SPDX-FileCopyrightText: Andreas Mueller
 # SPDX-FileCopyrightText: Joris Van den Bossche
-# SPDX-FileCopyrightText: Copyright (c) 2020-2025, NVIDIA CORPORATION.
+# SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
 # SPDX-License-Identifier: BSD-3-Clause
 
 # Original authors from Sckit-Learn:
@@ -36,6 +36,7 @@
 import cuml
 from cuml.internals.array_sparse import SparseCumlArray
 from cuml.internals.global_settings import _global_settings_data
+from cuml.internals.validation import check_is_fitted
 
 from ....thirdparty_adapters import check_array
 from ..preprocessing._function_transformer import FunctionTransformer
@@ -44,7 +45,6 @@
     BaseEstimator,
     TransformerMixin,
 )
-from ..utils.validation import check_is_fitted
 
 _ERR_MSG_1DCOLUMN = ("1D data passed to a transformer that expects 2D data. "
                      "Try to specify the column selection as a list of one "

python/cuml/cuml/_thirdparty/sklearn/preprocessing/_data.py

@@ -41,6 +41,7 @@
     StatelessTagMixin,
 )
 from cuml.internals.interop import InteropMixin, to_cpu, to_gpu
+from cuml.internals.validation import check_is_fitted
 
 from ....common.array_descriptor import CumlArrayDescriptor
 from ....internals.array import CumlArray
@@ -59,11 +60,7 @@
     mean_variance_axis,
     min_max_axis,
 )
-from ..utils.validation import (
-    FLOAT_DTYPES,
-    check_is_fitted,
-    check_random_state,
-)
+from ..utils.validation import FLOAT_DTYPES, check_random_state
 
 BOUNDS_THRESHOLD = 1e-7
 

python/cuml/cuml/_thirdparty/sklearn/preprocessing/_discretization.py

@@ -24,13 +24,14 @@
 from cuml.cluster import KMeans
 from cuml.internals.mixins import SparseInputTagMixin
 from cuml.preprocessing.encoders import OneHotEncoder
+from cuml.internals.validation import check_is_fitted
 
 from ....common.array_descriptor import CumlArrayDescriptor
 from ....internals.array_sparse import SparseCumlArray
 from ....internals.outputs import using_output_type, reflect
 from ....thirdparty_adapters import check_array
 from ..utils.skl_dependencies import BaseEstimator, TransformerMixin
-from ..utils.validation import FLOAT_DTYPES, check_is_fitted
+from ..utils.validation import FLOAT_DTYPES
 
 
 def digitize(x, bins):

python/cuml/cuml/_thirdparty/sklearn/preprocessing/_imputation.py

@@ -1,6 +1,6 @@
 # SPDX-FileCopyrightText: Nicolas Tresegnie <[email protected]>
 # SPDX-FileCopyrightText: Sergey Feldman <[email protected]>
-# SPDX-FileCopyrightText: Copyright (c) 2020-2025, NVIDIA CORPORATION.
+# SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
 # SPDX-License-Identifier: BSD-3-Clause
 
 # Original authors from Sckit-Learn:
@@ -28,6 +28,7 @@
     SparseInputTagMixin,
     StringInputTagMixin,
 )
+from cuml.internals.validation import check_is_fitted
 
 from ....common.array_descriptor import CumlArrayDescriptor
 from ....internals.array_sparse import SparseCumlArray
@@ -39,7 +40,7 @@
     _masked_column_mode,
 )
 from ..utils.skl_dependencies import BaseEstimator, TransformerMixin
-from ..utils.validation import FLOAT_DTYPES, check_is_fitted
+from ..utils.validation import FLOAT_DTYPES
 
 
 def is_scalar_nan(x):

python/cuml/cuml/_thirdparty/sklearn/utils/validation.py

@@ -5,7 +5,7 @@
 # SPDX-FileCopyrightText: Alexandre Gramfort
 # SPDX-FileCopyrightText: Nicolas Tresegnie
 # SPDX-FileCopyrightText: Sylvain Marie
-# SPDX-FileCopyrightText: Copyright (c) 2020-2025, NVIDIA CORPORATION.
+# SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
 # SPDX-License-Identifier: BSD-3-Clause
 
 # Original authors from Sckit-Learn:
@@ -32,7 +32,6 @@
 import cupyx.scipy.sparse as sp
 import numpy as np
 
-from ....common.exceptions import NotFittedError
 from ....thirdparty_adapters import check_array
 
 FLOAT_DTYPES = (np.float64, np.float32, np.float16)
@@ -176,74 +175,6 @@ def check_random_state(seed):
                      ' instance' % seed)
 
 
-def check_is_fitted(estimator, attributes=None, *, msg=None, all_or_any=all):
-    """Perform is_fitted validation for estimator.
-
-    Checks if the estimator is fitted by verifying the presence of
-    fitted attributes (ending with a trailing underscore) and otherwise
-    raises a NotFittedError with the given message.
-
-    This utility is meant to be used internally by estimators themselves,
-    typically in their own predict / transform methods.
-
-    Parameters
-    ----------
-    estimator : estimator instance.
-        estimator instance for which the check is performed.
-
-    attributes : str, list or tuple of str, default=None
-        Attribute name(s) given as string or a list/tuple of strings
-        Eg.: ``["coef_", "estimator_", ...], "coef_"``
-
-        If `None`, `estimator` is considered fitted if there exist an
-        attribute that ends with a underscore and does not start with double
-        underscore.
-
-    msg : string
-        The default error message is, "This %(name)s instance is not fitted
-        yet. Call 'fit' with appropriate arguments before using this
-        estimator."
-
-        For custom messages if "%(name)s" is present in the message string,
-        it is substituted for the estimator name.
-
-        Eg. : "Estimator, %(name)s, must be fitted before sparsifying".
-
-    all_or_any : callable, {all, any}, default all
-        Specify whether all or any of the given attributes must exist.
-
-    Returns
-    -------
-    None
-
-    Raises
-    ------
-    NotFittedError
-        If the attributes are not found.
-    """
-    if isclass(estimator):
-        raise TypeError("{} is a class, not an instance.".format(estimator))
-    if msg is None:
-        msg = ("This %(name)s instance is not fitted yet. Call 'fit' with "
-               "appropriate arguments before using this estimator.")
-
-    if not hasattr(estimator, 'fit'):
-        raise TypeError("%s is not an estimator instance." % (estimator))
-
-    if attributes is not None:
-        if not isinstance(attributes, (list, tuple)):
-            attributes = [attributes]
-        attrs = all_or_any([hasattr(estimator, attr) for attr in attributes])
-    elif hasattr(estimator, "__sklearn_is_fitted__"):
-        attrs = estimator.__sklearn_is_fitted__()
-    else:
-        attrs = [v for v in vars(estimator)
-                 if v.endswith("_") and not v.startswith("__")]
-
-    if not attrs:
-        raise NotFittedError(msg % {'name': type(estimator).__name__})
-
-
 def _allclose_dense_sparse(x, y, rtol=1e-7, atol=1e-9):
     """Check allclose for sparse and dense data.
 

python/cuml/cuml/cluster/hdbscan/hdbscan.pyx

@@ -18,6 +18,7 @@ from cuml.internals.interop import (
 )
 from cuml.internals.mem_type import MemoryType
 from cuml.internals.mixins import ClusterMixin, CMajorInputTagMixin
+from cuml.internals.validation import check_is_fitted
 
 from cython.operator cimport dereference as deref
 from libc.stdint cimport int64_t, uint64_t, uintptr_t
@@ -906,8 +907,7 @@ class HDBSCAN(Base, InteropMixin, ClusterMixin, CMajorInputTagMixin):
         the label of new/unseen points. This data is only useful if you
         are intending to use functions from hdbscan.prediction.
         """
-        if getattr(self, "labels_", None) is None:
-            raise ValueError("The model is not trained yet (call fit() first).")
+        check_is_fitted(self)
 
         with cuml.using_output_type("cuml"):
             labels = self.labels_
@@ -1107,10 +1107,8 @@ def _check_clusterer(clusterer):
             f"Expected an instance of `HDBSCAN`, got {type(clusterer).__name__}"
         )
 
-    if getattr(clusterer, "labels_", None) is None:
-        raise ValueError(
-            "The clusterer is not fit, please call `clusterer.fit` first"
-        )
+    check_is_fitted(clusterer)
+
     cdef _HDBSCANState state = <_HDBSCANState?>clusterer._state
 
     if state.prediction_data == NULL:

python/cuml/cuml/cluster/kmeans.pyx

@@ -19,7 +19,7 @@ from cuml.internals.interop import (
 )
 from cuml.internals.mixins import ClusterMixin, CMajorInputTagMixin
 from cuml.internals.outputs import reflect, run_in_internal_context
-from cuml.internals.utils import check_random_seed
+from cuml.internals.validation import check_is_fitted, check_random_seed
 
 from libc.stdint cimport int64_t, uintptr_t
 from libcpp cimport bool
@@ -633,6 +633,8 @@ class KMeans(Base,
         inertia : float
         Sum of squared distances of samples to their closest cluster center.
         """
+        check_is_fitted(self)
+
         dtype = self.cluster_centers_.dtype
 
         X_m, n_rows, _, _ = input_to_cuml_array(
@@ -694,6 +696,8 @@ class KMeans(Base,
         Transform X to a cluster-distance space.
 
         """
+        check_is_fitted(self)
+
         dtype = self.cluster_centers_.dtype
 
         X_m = input_to_cuml_array(

python/cuml/cuml/cluster/spectral_clustering.pyx

@@ -14,7 +14,7 @@ from cuml.common.array_descriptor import CumlArrayDescriptor
 from cuml.internals.array import CumlArray
 from cuml.internals.base import Base, get_handle
 from cuml.internals.input_utils import input_to_cupy_array
-from cuml.internals.utils import check_random_seed
+from cuml.internals.validation import check_random_seed
 
 from libc.stdint cimport uint64_t, uintptr_t
 from libcpp cimport bool

python/cuml/cuml/common/exceptions.py

@@ -1,12 +1,23 @@
 #
-# SPDX-FileCopyrightText: Copyright (c) 2020, NVIDIA CORPORATION.
+# SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
 # SPDX-License-Identifier: Apache-2.0
 #
+__all__ = ("NotFittedError",)  # noqa
 
 
-class NotFittedError(ValueError, AttributeError):
-    """Exception class to raise if estimator is used before fitting.
+def __getattr__(name):
+    if name == "NotFittedError":
+        import warnings
 
-    This class inherits from both ValueError and AttributeError to help with
-    exception handling and backward compatibility.
-    """
+        from sklearn.exceptions import NotFittedError
+
+        warnings.warn(
+            "`cuml.common.exceptions.NotFittedError` was deprecated in 26.04 "
+            "and will be removed in 26.06. Please use "
+            "`sklearn.exceptions.NotFittedError` instead.",
+            FutureWarning,
+            stacklevel=2,
+        )
+        return NotFittedError
+    else:
+        raise AttributeError(f"module {__name__} has no attribute {name}")

python/cuml/cuml/covariance/ledoit_wolf.py

@@ -14,6 +14,7 @@
 from cuml.internals.base import Base
 from cuml.internals.input_utils import input_to_cupy_array
 from cuml.internals.interop import InteropMixin, to_cpu, to_gpu
+from cuml.internals.validation import check_is_fitted
 
 
 def _ledoit_wolf_shrinkage(X, assume_centered=False, block_size=1000):
@@ -299,6 +300,8 @@ def get_precision(self):
         precision_ : ndarray of shape (n_features, n_features)
             The precision matrix associated to the current covariance object.
         """
+        check_is_fitted(self)
+
         if self.store_precision:
             return self.precision_
         else:
@@ -324,6 +327,8 @@ def score(self, X_test, y=None) -> float:
         log_likelihood : float
             Log-likelihood of the data under the fitted Gaussian model.
         """
+        check_is_fitted(self)
+
         X_arr, _, n_features, _ = input_to_cupy_array(
             X_test,
             check_dtype=[np.float32, np.float64],
@@ -367,6 +372,8 @@ def error_norm(
             The Mean Squared Error (in the sense of the Frobenius norm)
             between `self` and `comp_cov`.
         """
+        check_is_fitted(self)
+
         comp_cov_arr, _, _, _ = input_to_cupy_array(
             comp_cov,
             check_dtype=[np.float32, np.float64],
@@ -408,6 +415,8 @@ def mahalanobis(self, X):
         mahalanobis_distances : ndarray of shape (n_samples,)
             Squared Mahalanobis distances of the observations.
         """
+        check_is_fitted(self)
+
         X_arr, _, _, _ = input_to_cupy_array(
             X,
             check_dtype=[np.float32, np.float64],

python/cuml/cuml/dask/cluster/kmeans.py

@@ -14,7 +14,7 @@
 )
 from cuml.dask.common.input_utils import DistributedDataHandler, concatenate
 from cuml.dask.common.utils import wait_and_raise_from_futures
-from cuml.internals.utils import check_random_seed
+from cuml.internals.validation import check_random_seed
 
 
 class KMeans(BaseEstimator, DelayedPredictionMixin, DelayedTransformMixin):

python/cuml/cuml/dask/preprocessing/LabelEncoder.py

@@ -1,13 +1,13 @@
-# SPDX-FileCopyrightText: Copyright (c) 2021-2025, NVIDIA CORPORATION.
+# SPDX-FileCopyrightText: Copyright (c) 2021-2026, NVIDIA CORPORATION.
 # SPDX-License-Identifier: Apache-2.0
 #
 from collections.abc import Sequence
 
 from dask_cudf import DataFrame as dcDataFrame
 from dask_cudf import Series as dcSeries
+from sklearn.exceptions import NotFittedError
 from toolz import first
 
-from cuml.common.exceptions import NotFittedError
 from cuml.dask.common.base import (
     BaseEstimator,
     DelayedInverseTransformMixin,

python/cuml/cuml/datasets/classification.py

@@ -8,7 +8,7 @@
 import cuml.internals
 import cuml.internals.nvtx as nvtx
 from cuml.datasets.utils import _create_rs_generator
-from cuml.internals.utils import check_random_seed
+from cuml.internals.validation import check_random_seed
 
 
 def _generate_hypercube(samples, dimensions, random_state):

python/cuml/cuml/decomposition/incremental_pca.py

@@ -15,6 +15,7 @@
 from cuml.internals.array import CumlArray
 from cuml.internals.base import Base
 from cuml.internals.input_utils import input_to_cupy_array
+from cuml.internals.validation import check_is_fitted
 
 
 class IncrementalPCA(PCA):
@@ -418,6 +419,7 @@ def transform(self, X, *, convert_dtype=False) -> CumlArray:
         X_new : array-like, shape (n_samples, n_components)
 
         """
+        check_is_fitted(self)
 
         if scipy.sparse.issparse(X) or cupyx.scipy.sparse.issparse(X):
             X = _validate_sparse_input(X)