Add sklearnex interface for polynomial kernel

sklearnex/svm/poly_kernel.py

@@ -0,0 +1,74 @@
+# ==============================================================================
+# Copyright contributors to the oneDAL project
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ==============================================================================
+
+import numpy as np
+from scipy import sparse
+from sklearn.metrics.pairwise import polynomial_kernel as sklearn_poly_kernel
+
+from onedal.primitives import poly_kernel as onedal_poly_kernel
+
+
+def poly_kernel(
+    X,
+    Y=None,
+    degree=3,
+    gamma=None,
+    coef0=1,
+    queue=None,
+):
+    """
+    Compute the polynomial kernel using the oneDAL backend when possible.
+
+    Falls back to scikit-learn's ``polynomial_kernel`` for unsupported cases
+    such as sparse inputs.
+
+    Parameters
+    ----------
+    X : array-like of shape (n_samples_X, n_features)
+        Input feature array.
+
+    Y : array-like of shape (n_samples_Y, n_features), default=None
+        Optional second feature array. If None, ``Y = X``.
+
+    degree : int, default=3
+        Degree of the polynomial kernel.
+
+    gamma : float, default=None
+        Scaling factor for the inner product. If None, ``1 / n_features`` is used.
+
+    coef0 : float, default=1
+        Constant term added to the scaled inner product.
+
+    queue : SyclQueue or None, default=None
+        Optional SYCL queue for device execution.
+
+    Returns
+    -------
+    kernel_matrix : ndarray of shape (n_samples_X, n_samples_Y)
+        Computed polynomial kernel Gram matrix.
+    """
+    # Fall back to sklearn if sparse input
+    if sparse.issparse(X) or (Y is not None and sparse.issparse(Y)):
+        return sklearn_poly_kernel(X, Y, degree=degree, gamma=gamma, coef0=coef0)
+
+    # Handle gamma default like sklearn
+    if gamma is None:
+        gamma = 1.0 / X.shape[1]
+
+    # Use oneDAL accelerated path
+    return onedal_poly_kernel(
+        X, Y=Y, gamma=gamma, coef0=coef0, degree=degree, queue=queue
+    )

sklearnex/svm/tests/test_poly_kernel_sklearnex.py

@@ -0,0 +1,43 @@
+# ==============================================================================
+# Copyright contributors to the oneDAL project
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ==============================================================================
+
+import numpy as np
+from scipy import sparse
+from sklearn.metrics.pairwise import polynomial_kernel as sklearn_poly_kernel
+
+from sklearnex.svm.poly_kernel import poly_kernel
+
+
+def test_poly_kernel_dense():
+    """Test poly_kernel on dense input arrays."""
+    X = np.array([[1, 2], [3, 4]])
+    Y = np.array([[5, 6], [7, 8]])
+
+    result = poly_kernel(X, Y, degree=2, gamma=0.5, coef0=1)
+    expected = sklearn_poly_kernel(X, Y, degree=2, gamma=0.5, coef0=1)
+
+    np.testing.assert_allclose(result, expected, rtol=1e-5, atol=1e-8)
+
+
+def test_poly_kernel_sparse_fallback():
+    """Test that poly_kernel falls back to sklearn implementation for sparse inputs."""
+    X = sparse.csr_matrix([[1, 0], [0, 1]])
+    Y = sparse.csr_matrix([[0, 1], [1, 0]])
+
+    result = poly_kernel(X, Y, degree=3, gamma=1.0, coef0=1.0)
+    expected = sklearn_poly_kernel(X, Y, degree=3, gamma=1.0, coef0=1.0)
+
+    np.testing.assert_allclose(result, expected, rtol=1e-5, atol=1e-8)