TEST: using XFAIL instead of SKIP where it is required

daal4py/sklearn/monkeypatch/tests/test_patching.py

@@ -86,5 +86,7 @@ def test_patching(configuration):
         return
     for skip in TO_SKIP:
         if re.search(skip, configuration) is not None:
+            # TODO:
+            # check if xfail needed instead.
             pytest.skip("SKIPPED", allow_module_level=False)
     raise ValueError("Test patching failed: " + configuration)

onedal/basic_statistics/tests/test_basic_statistics.py

@@ -214,7 +214,7 @@ def test_options_csr(queue, option, dtype):
     fp32tol, fp64tol = tols
 
     if result_option == "max":
-        pytest.skip("There is a bug in oneDAL's max computations on GPU")
+        pytest.xfail("There is a bug in oneDAL's max computations on GPU")
 
     seed = 42
     row_count, column_count = 20046, 4007

onedal/common/tests/test_policy.py

@@ -49,7 +49,7 @@ def test_with_numpy_data(queue):
 @pytest.mark.parametrize("memtype", get_memory_usm())
 def test_with_usm_ndarray_data(queue, memtype):
     if queue is None:
-        pytest.skip(
+        pytest.xfail(
             "dpctl Memory object with queue=None uses cached default (gpu if available)"
         )
 

onedal/datatypes/tests/test_data.py

@@ -108,7 +108,7 @@ def _test_input_format_c_contiguous_numpy(queue, dtype):
 @pytest.mark.parametrize("dtype", [np.float32, np.float64])
 def test_input_format_c_contiguous_numpy(queue, dtype):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Sporadic failures on GPU sycl_queue.")
+        pytest.xfail("Sporadic failures on GPU sycl_queue.")
     _test_input_format_c_contiguous_numpy(queue, dtype)
 
 
@@ -130,7 +130,7 @@ def _test_input_format_f_contiguous_numpy(queue, dtype):
 @pytest.mark.parametrize("dtype", [np.float32, np.float64])
 def test_input_format_f_contiguous_numpy(queue, dtype):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Sporadic failures on GPU sycl_queue.")
+        pytest.xfail("Sporadic failures on GPU sycl_queue.")
     _test_input_format_f_contiguous_numpy(queue, dtype)
 
 
@@ -156,7 +156,7 @@ def _test_input_format_c_not_contiguous_numpy(queue, dtype):
 @pytest.mark.parametrize("dtype", [np.float32, np.float64])
 def test_input_format_c_not_contiguous_numpy(queue, dtype):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Sporadic failures on GPU sycl_queue.")
+        pytest.xfail("Sporadic failures on GPU sycl_queue.")
     _test_input_format_c_not_contiguous_numpy(queue, dtype)
 
 
@@ -180,7 +180,7 @@ def _test_input_format_c_contiguous_pandas(queue, dtype):
 @pytest.mark.parametrize("dtype", [np.float32, np.float64])
 def test_input_format_c_contiguous_pandas(queue, dtype):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Sporadic failures on GPU sycl_queue.")
+        pytest.xfail("Sporadic failures on GPU sycl_queue.")
     _test_input_format_c_contiguous_pandas(queue, dtype)
 
 
@@ -204,7 +204,7 @@ def _test_input_format_f_contiguous_pandas(queue, dtype):
 @pytest.mark.parametrize("dtype", [np.float32, np.float64])
 def test_input_format_f_contiguous_pandas(queue, dtype):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Sporadic failures on GPU sycl_queue.")
+        pytest.xfail("Sporadic failures on GPU sycl_queue.")
     _test_input_format_f_contiguous_pandas(queue, dtype)
 
 

onedal/ensemble/tests/test_random_forest.py

@@ -40,7 +40,7 @@ def test_rf_classifier(queue):
 @pytest.mark.parametrize("queue", get_queues())
 def test_rf_regression(queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("RF regressor predict for the GPU sycl_queue is buggy.")
+        pytest.xfail("RF regressor predict for the GPU sycl_queue is buggy.")
     X, y = make_regression(
         n_samples=100, n_features=4, n_informative=2, random_state=0, shuffle=False
     )
@@ -84,7 +84,7 @@ def test_rf_regression_random_splitter(queue):
     # splitter_mode selection only for GPU enabled.
     # For CPU only `best` mode is supported.
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("RF regressor predict for the GPU sycl_queue is buggy.")
+        pytest.xfail("RF regressor predict for the GPU sycl_queue is buggy.")
     X, y = make_regression(
         n_samples=100, n_features=4, n_informative=2, random_state=0, shuffle=False
     )

onedal/neighbors/tests/test_knn_classification.py

@@ -34,7 +34,7 @@ def test_iris(queue):
 @pytest.mark.parametrize("queue", get_queues())
 def test_pickle(queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("KNN classifier pickling for the GPU sycl_queue is buggy.")
+        pytest.xfail("KNN classifier pickling for the GPU sycl_queue is buggy.")
     iris = datasets.load_iris()
     clf = KNeighborsClassifier(2).fit(iris.data, iris.target, queue=queue)
     expected = clf.predict(iris.data, queue=queue)

onedal/primitives/tests/test_kernel_functions.py

@@ -29,7 +29,7 @@
 @pytest.mark.parametrize("queue", get_queues())
 def test_dense_self_linear_kernel(queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Linear kernel for the GPU sycl_queue is buggy.")
+        pytest.xfail("Linear kernel for the GPU sycl_queue is buggy.")
     rng = np.random.RandomState(0)
     X = np.array(5 * rng.random_sample((10, 4)))
 
@@ -56,7 +56,7 @@ def _test_dense_small_linear_kernel(queue, scale, shift, dtype):
 @pytest.mark.parametrize("dtype", [np.float32, np.float64])
 def test_dense_small_linear_kernel(queue, scale, shift, dtype):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Linear kernel for the GPU sycl_queue is buggy.")
+        pytest.xfail("Linear kernel for the GPU sycl_queue is buggy.")
     _test_dense_small_linear_kernel(queue, scale, shift, dtype)
 
 

onedal/svm/tests/test_csr_svm.py

@@ -143,7 +143,7 @@ def _test_iris(queue, kernel):
 @pytest.mark.parametrize("kernel", ["linear", "rbf", "poly", "sigmoid"])
 def test_iris(queue, kernel):
     if kernel == "rbf":
-        pytest.skip("RBF CSR SVM test failing in 2025.0.")
+        pytest.xfail("RBF CSR SVM test failing in 2025.0.")
     _test_iris(queue, kernel)
 
 
@@ -163,7 +163,7 @@ def _test_diabetes(queue, kernel):
 @pytest.mark.parametrize("kernel", ["linear", "rbf", "poly", "sigmoid"])
 def test_diabetes(queue, kernel):
     if kernel == "sigmoid":
-        pytest.skip("Sparse sigmoid kernel function is buggy.")
+        pytest.xfail("Sparse sigmoid kernel function is buggy.")
     _test_diabetes(queue, kernel)
 
 

onedal/svm/tests/test_nusvr.py

@@ -110,7 +110,7 @@ def _test_diabetes_compare_with_sklearn(queue, kernel):
 @pytest.mark.parametrize("kernel", ["linear", "rbf", "poly", "sigmoid"])
 def test_diabetes_compare_with_sklearn(queue, kernel):
     if kernel == "sigmoid":
-        pytest.skip("Sparse sigmoid kernel function is buggy.")
+        pytest.xfail("Sparse sigmoid kernel function is buggy.")
     _test_diabetes_compare_with_sklearn(queue, kernel)
 
 

onedal/svm/tests/test_svc.py

@@ -47,7 +47,7 @@ def _test_libsvm_parameters(queue, array_constr, dtype):
 @pytest.mark.parametrize("dtype", [np.float32, np.float64])
 def test_libsvm_parameters(queue, array_constr, dtype):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Sporadic failures on GPU sycl_queue.")
+        pytest.xfail("Sporadic failures on GPU sycl_queue.")
     _test_libsvm_parameters(queue, array_constr, dtype)
 
 
@@ -76,7 +76,7 @@ def test_class_weight(queue):
 @pytest.mark.parametrize("queue", get_queues())
 def test_sample_weight(queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("Sporadic failures on GPU sycl_queue.")
+        pytest.xfail("Sporadic failures on GPU sycl_queue.")
     X = np.array([[-2, 0], [-1, -1], [0, -2], [0, 2], [1, 1], [2, 2]])
     y = np.array([1, 1, 1, 2, 2, 2])
 

onedal/svm/tests/test_svr.py

@@ -125,7 +125,7 @@ def _test_diabetes_compare_with_sklearn(queue, kernel):
 @pytest.mark.parametrize("kernel", ["linear", "rbf", "poly", "sigmoid"])
 def test_diabetes_compare_with_sklearn(queue, kernel):
     if kernel == "sigmoid":
-        pytest.skip("Sparse sigmoid kernel function is buggy.")
+        pytest.xfail("Sparse sigmoid kernel function is buggy.")
     _test_diabetes_compare_with_sklearn(queue, kernel)
 
 

sklearnex/covariance/tests/test_incremental_covariance.py

@@ -40,7 +40,7 @@
 def test_sklearnex_partial_fit_on_gold_data(dataframe, queue, dtype, assume_centered):
     is_gpu = queue is not None and queue.sycl_device.is_gpu
     if assume_centered and is_gpu and not daal_check_version((2025, "P", 0)):
-        pytest.skip(
+        pytest.xfail(
             "Due to a bug on oneDAL side, means are not set to zero when assume_centered=True"
         )
     from sklearnex.covariance import IncrementalEmpiricalCovariance
@@ -151,7 +151,7 @@ def test_sklearnex_fit_on_random_data(
 ):
     is_gpu = queue is not None and queue.sycl_device.is_gpu
     if assume_centered and is_gpu and not daal_check_version((2025, "P", 0)):
-        pytest.skip(
+        pytest.xfail(
             "Due to a bug on oneDAL side, means are not set to zero when assume_centered=True"
         )
     from sklearnex.covariance import IncrementalEmpiricalCovariance

sklearnex/ensemble/tests/test_forest.py

@@ -64,7 +64,7 @@ def test_sklearnex_import_rf_classifier(dataframe, queue, block, trees, rows, sc
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 def test_sklearnex_import_rf_regression(dataframe, queue):
     if (not daal_check_version((2025, "P", 200))) and queue and queue.sycl_device.is_gpu:
-        pytest.skip("Skipping due to bug in histogram merges fixed in 2025.2.")
+        pytest.xfail("Skipping due to bug in histogram merges fixed in 2025.2.")
     from sklearnex.ensemble import RandomForestRegressor
 
     X, y = make_regression(n_features=4, n_informative=2, random_state=0, shuffle=False)
@@ -87,7 +87,7 @@ def test_sklearnex_import_rf_regression(dataframe, queue):
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 def test_sklearnex_import_et_classifier(dataframe, queue):
     if (not daal_check_version((2025, "P", 200))) and queue and queue.sycl_device.is_gpu:
-        pytest.skip("Skipping due to bug in histogram merges fixed in 2025.2.")
+        pytest.xfail("Skipping due to bug in histogram merges fixed in 2025.2.")
     from sklearnex.ensemble import ExtraTreesClassifier
 
     X, y = make_classification(
@@ -110,7 +110,7 @@ def test_sklearnex_import_et_classifier(dataframe, queue):
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 def test_sklearnex_import_et_regression(dataframe, queue):
     if (not daal_check_version((2025, "P", 200))) and queue and queue.sycl_device.is_gpu:
-        pytest.skip("Skipping due to bug in histogram merges fixed in 2025.2.")
+        pytest.xfail("Skipping due to bug in histogram merges fixed in 2025.2.")
     from sklearnex.ensemble import ExtraTreesRegressor
 
     X, y = make_regression(n_features=1, random_state=0, shuffle=False)

sklearnex/spmd/decomposition/tests/test_pca_spmd.py

@@ -96,6 +96,8 @@ def test_pca_spmd_synthetic(
     n_samples, n_features, n_components, whiten, dataframe, queue, dtype
 ):
     # TODO: Resolve issues with batch fallback and lack of support for n_rows_rank < n_cols
+    # TODO:
+    # check if xfail is needed instead.
     if n_components == "mle" or n_components == 3:
         pytest.skip("Avoid error in case of batch fallback to sklearn")
     if n_samples <= n_features:

sklearnex/spmd/ensemble/tests/test_forest_spmd.py

@@ -91,7 +91,7 @@ def test_rfcls_spmd_gold(dataframe, queue):
     spmd_result = spmd_model.predict(local_dpt_X_test)
     batch_result = batch_model.predict(X_test)
 
-    pytest.skip("SPMD and batch random forest results not aligned")
+    pytest.xfail("SPMD and batch random forest results not aligned")
     _spmd_assert_allclose(spmd_result, batch_result)
 
 
@@ -144,7 +144,7 @@ def test_rfcls_spmd_synthetic(
     spmd_result = spmd_model.predict(local_dpt_X_test)
     batch_result = batch_model.predict(X_test)
 
-    pytest.skip("SPMD and batch random forest results not aligned")
+    pytest.xfail("SPMD and batch random forest results not aligned")
     _spmd_assert_allclose(spmd_result, batch_result)
 
 
@@ -208,7 +208,7 @@ def test_rfreg_spmd_gold(dataframe, queue):
     spmd_result = spmd_model.predict(local_dpt_X_test)
     batch_result = batch_model.predict(X_test)
 
-    pytest.skip("SPMD and batch random forest results not aligned")
+    pytest.xfail("SPMD and batch random forest results not aligned")
     _spmd_assert_allclose(spmd_result, batch_result)
 
 
@@ -261,5 +261,5 @@ def test_rfreg_spmd_synthetic(
     batch_result = batch_model.predict(X_test)
 
     # TODO: remove skips when SPMD and batch are aligned
-    pytest.skip("SPMD and batch random forest results not aligned")
+    pytest.xfail("SPMD and batch random forest results not aligned")
     _spmd_assert_allclose(spmd_result, batch_result)

sklearnex/spmd/linear_model/tests/test_logistic_regression_spmd.py

@@ -118,7 +118,7 @@ def test_logistic_spmd_gold(dataframe, queue):
 def test_logistic_spmd_synthetic(n_samples, n_features, C, tol, dataframe, queue, dtype):
     # TODO: Resolve numerical issues when n_rows_rank < n_cols
     if n_samples <= n_features:
-        pytest.skip("Numerical issues when rank rows < columns")
+        pytest.xfail("Numerical issues when rank rows < columns")
 
     # Import spmd and batch algo
     from sklearnex.linear_model import LogisticRegression as LogisticRegression_Batch

sklearnex/svm/tests/test_svm.py

@@ -28,7 +28,7 @@
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 def test_sklearnex_import_svc(dataframe, queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("SVC fit for the GPU sycl_queue is buggy.")
+        pytest.xfail("SVC fit for the GPU sycl_queue is buggy.")
     from sklearnex.svm import SVC
 
     X = np.array([[-2, -1], [-1, -1], [-1, -2], [+1, +1], [+1, +2], [+2, +1]])
@@ -44,7 +44,7 @@ def test_sklearnex_import_svc(dataframe, queue):
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 def test_sklearnex_import_nusvc(dataframe, queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("NuSVC fit for the GPU sycl_queue is buggy.")
+        pytest.xfail("NuSVC fit for the GPU sycl_queue is buggy.")
     from sklearnex.svm import NuSVC
 
     X = np.array([[-2, -1], [-1, -1], [-1, -2], [+1, +1], [+1, +2], [+2, +1]])
@@ -62,7 +62,7 @@ def test_sklearnex_import_nusvc(dataframe, queue):
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 def test_sklearnex_import_svr(dataframe, queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("SVR fit for the GPU sycl_queue is buggy.")
+        pytest.xfail("SVR fit for the GPU sycl_queue is buggy.")
     from sklearnex.svm import SVR
 
     X = np.array([[-2, -1], [-1, -1], [-1, -2], [+1, +1], [+1, +2], [+2, +1]])
@@ -78,7 +78,7 @@ def test_sklearnex_import_svr(dataframe, queue):
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 def test_sklearnex_import_nusvr(dataframe, queue):
     if queue and queue.sycl_device.is_gpu:
-        pytest.skip("NuSVR fit for the GPU sycl_queue is buggy.")
+        pytest.xfail("NuSVR fit for the GPU sycl_queue is buggy.")
     from sklearnex.svm import NuSVR
 
     X = np.array([[-2, -1], [-1, -1], [-1, -2], [+1, +1], [+1, +2], [+2, +1]])

sklearnex/tests/test_run_to_run_stability.py

@@ -162,7 +162,7 @@ def test_standard_estimator_stability(estimator, method, dataframe, queue):
     if estimator in ["LogisticRegression", "TSNE"]:
         pytest.skip(f"stability not guaranteed for {estimator}")
     if estimator in ["KMeans", "PCA"] and "score" in method and queue == None:
-        pytest.skip(f"variation observed in {estimator}.score")
+        pytest.xfail(f"variation observed in {estimator}.score")
     if estimator in ["IncrementalEmpiricalCovariance"] and method == "mahalanobis":
         pytest.skip("allowed fallback to sklearn occurs")
     _skip_neighbors(estimator, method)
@@ -191,9 +191,9 @@ def test_special_estimator_stability(estimator, method, dataframe, queue):
     if queue is None and estimator in ["LogisticRegression(solver='newton-cg')"]:
         pytest.skip(f"stability not guaranteed for {estimator}")
     if "KMeans" in estimator and method == "score" and queue == None:
-        pytest.skip(f"variation observed in KMeans.score")
+        pytest.xfail(f"variation observed in KMeans.score")
     if estimator == "BasicStatistics()" and queue == None:
-        pytest.skip(f"BasicStatistics not deterministic")
+        pytest.xfail(f"BasicStatistics not deterministic")
     if "NearestNeighbors" in estimator and "radius" in method:
         pytest.skip(f"RadiusNeighbors estimator not implemented in sklearnex")
     _skip_neighbors(estimator, method)
@@ -216,13 +216,13 @@ def test_special_estimator_stability(estimator, method, dataframe, queue):
 @pytest.mark.parametrize("estimator, method", gen_models_info(SPARSE_INSTANCES))
 def test_sparse_estimator_stability(estimator, method, dataframe, queue):
     if "KMeans" in estimator and method in "score" and queue == None:
-        pytest.skip(f"variation observed in KMeans.{method}")
+        pytest.xfail(f"variation observed in KMeans.{method}")
     if (
         not daal_check_version((2025, "P", 0))
         and "KMeans()" in estimator
         and queue == None
     ):
-        pytest.skip(f"variation observed in KMeans.{method} in 2024.7 oneDAL")
+        pytest.xfail(f"variation observed in KMeans.{method} in 2024.7 oneDAL")
     if "NearestNeighbors" in estimator and "radius" in method:
         pytest.skip(f"RadiusNeighbors estimator not implemented in sklearnex")
     _skip_neighbors(estimator, method)
@@ -247,7 +247,7 @@ def test_sparse_estimator_stability(estimator, method, dataframe, queue):
 @pytest.mark.parametrize("estimator, method", gen_models_info(STABILITY_INSTANCES))
 def test_other_estimator_stability(estimator, method, dataframe, queue):
     if "KMeans" in estimator and method == "score" and queue == None:
-        pytest.skip(f"variation observed in KMeans.score")
+        pytest.xfail(f"variation observed in KMeans.score")
     if "NearestNeighbors" in estimator and "radius" in method:
         pytest.skip(f"RadiusNeighbors estimator not implemented in sklearnex")
     _skip_neighbors(estimator, method)