Merge branch 'master' into enable-clip-tests

Author: antonwolfy

.github/workflows/build-sphinx.yml

@@ -42,6 +42,7 @@ jobs:
       oneapi-pkgs-env: ''
       # Enable env when it's required to use only conda packages without OneAPI installation
       # oneapi-pkgs-env: '${{ github.workspace }}/environments/oneapi_pkgs.yml'
+      dpctl-pkg-txt: 'environments/dpctl_pkg.txt'
 
     steps:
       - name: Cancel Previous Runs
@@ -149,7 +150,7 @@ jobs:
       - name: Install dpctl
         if: env.oneapi-pkgs-env == ''
         run: |
-          pip install -i https://pypi.anaconda.org/dppy/label/dev/simple dpctl==0.20.0dev0
+          pip install -r ${{ env.dpctl-pkg-txt }}
 
       - name: Conda info
         run: |

.github/workflows/check-mkl-interfaces.yaml

@@ -48,7 +48,7 @@ jobs:
 
       - name: Merge conda env files
         run: |
-          conda-merge ${{ env.build-with-oneapi-env }} ${{ env.dpctl-pkg-env }} ${{ env.oneapi-pkgs-env }} > ${{ env.environment-file }}
+          conda-merge ${{ env.dpctl-pkg-env }} ${{ env.oneapi-pkgs-env }} ${{ env.build-with-oneapi-env }} > ${{ env.environment-file }}
           cat ${{ env.environment-file }}
 
       - name: Upload artifact
@@ -229,7 +229,7 @@ jobs:
           python -c "import dpnp; print(dpnp.__version__)"
 
       - name: Run tests
-        if: env.rerun-tests-on-failure == 'true'
+        if: env.rerun-tests-on-failure != 'true'
         run: |
           python -m pytest -ra --pyargs dpnp.tests
         env:

.github/workflows/conda-package.yml

@@ -82,6 +82,9 @@ jobs:
           environment-file: ${{ env.build-conda-pkg-env }}
           activate-environment: ${{ env.build-env-name }}
 
+      - name: List installed packages
+        run: mamba list
+
       - name: Store conda paths as envs
         shell: bash -el {0}
         run: |
@@ -93,13 +96,13 @@ jobs:
         continue-on-error: true
         run: conda build --no-test --python ${{ matrix.python }} --numpy 2.0 ${{ env.channels-list }} conda-recipe
         env:
-          MAX_BUILD_CMPL_MKL_VERSION: '2025.1a0'
+          MAX_BUILD_CMPL_MKL_VERSION: '2025.2a0'
 
       - name: ReBuild conda package
         if: steps.build_conda_pkg.outcome == 'failure'
         run: conda build --no-test --python ${{ matrix.python }} --numpy 2.0 ${{ env.channels-list }} conda-recipe
         env:
-          MAX_BUILD_CMPL_MKL_VERSION: '2025.1a0'
+          MAX_BUILD_CMPL_MKL_VERSION: '2025.2a0'
 
       - name: Upload artifact
         uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
@@ -625,6 +628,7 @@ jobs:
           python -m pytest --json-report --json-report-file=${{ env.json-report-file }} --disable-deadline --skips-file ${{ env.array-api-skips-file }} array_api_tests || true
         env:
           ARRAY_API_TESTS_MODULE: 'dpnp'
+          ARRAY_API_TESTS_VERSION: '2024.12'
           SYCL_CACHE_PERSISTENT: 1
         working-directory: ${{ env.array-api-tests-path }}
 

.github/workflows/generate_coverage.yaml

@@ -26,6 +26,7 @@ jobs:
       oneapi-pkgs-env: ''
       # Enable env when it's required to use only conda packages without OneAPI installation
       # oneapi-pkgs-env: '${{ github.workspace }}/environments/oneapi_pkgs.yml'
+      dpctl-pkg-txt: 'environments/dpctl_pkg.txt'
 
     steps:
       - name: Cancel Previous Runs
@@ -108,7 +109,7 @@ jobs:
       - name: Install dpctl
         if: env.oneapi-pkgs-env == ''
         run: |
-          pip install -i https://pypi.anaconda.org/dppy/label/dev/simple dpctl==0.20.0dev0
+          pip install -r ${{ env.dpctl-pkg-txt }}
 
       - name: Conda info
         run: |

.github/workflows/openssf-scorecard.yml

@@ -68,6 +68,6 @@ jobs:
 
       # Upload the results to GitHub's code scanning dashboard.
       - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@5f8171a638ada777af81d42b55959a643bb29017 # v3.28.12
+        uses: github/codeql-action/upload-sarif@45775bd8235c68ba998cffa5171334d58593da47 # v3.28.15
         with:
           sarif_file: results.sarif

.github/workflows/pre-commit.yml

@@ -29,7 +29,7 @@ jobs:
         uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
 
       - name: Set up python
-        uses: actions/setup-python@42375524e23c412d93fb67b49958b491fce71c38 # v5.4.0
+        uses: actions/setup-python@8d9ed9ac5c53483de85588cdf95a591a75ab9f55 # v5.5.0
         with:
           python-version: '3.13'
 

.pre-commit-config.yaml

@@ -2,13 +2,13 @@
 # See https://pre-commit.com/hooks.html for more hooks
 repos:
 -   repo: https://github.com/PyCQA/bandit
-    rev: '1.7.9'
+    rev: '1.8.3'
     hooks:
     -   id: bandit
         pass_filenames: false
         args: ["-r", "dpnp", "-lll"]
 -   repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.6.0
+    rev: v5.0.0
     hooks:
     -   id: check-ast
     -   id: check-builtin-literals
@@ -44,17 +44,18 @@ repos:
     -   id: rst-inline-touching-normal
     -   id: text-unicode-replacement-char
 -   repo: https://github.com/codespell-project/codespell
-    rev: v2.3.0
+    rev: v2.4.1
     hooks:
     -   id: codespell
+        args: ["-L", "abd"]  # ignore "abd" used in einsum tests
         additional_dependencies:
             - tomli
 -   repo: https://github.com/psf/black
-    rev: 24.4.2
+    rev: 25.1.0
     hooks:
     -   id: black
 -   repo: https://github.com/pycqa/isort
-    rev: 5.13.2
+    rev: 6.0.1
     hooks:
     -   id: isort
         name: isort (python)
@@ -65,20 +66,20 @@ repos:
         name: isort (pyi)
         types: [pyi]
 -   repo: https://github.com/pycqa/flake8
-    rev: 7.1.0
+    rev: 7.1.2
     hooks:
     -   id: flake8
         args: ["--config=.flake8"]
         additional_dependencies:
             - flake8-docstrings==1.7.0
-            - flake8-bugbear==24.4.26
+            - flake8-bugbear==24.12.12
 -   repo: https://github.com/pocc/pre-commit-hooks
     rev: v1.3.5
     hooks:
     -   id: clang-format
         args: ["-i"]
 -   repo: https://github.com/gitleaks/gitleaks
-    rev: v8.18.4
+    rev: v8.24.0
     hooks:
     -   id: gitleaks
 -   repo: https://github.com/jumanjihouse/pre-commit-hooks
@@ -102,3 +103,8 @@ repos:
             "--disable=unused-wildcard-import"
             ]
         files: '^dpnp/(dpnp_iface.*|fft|linalg)'
+-   repo: https://github.com/macisamuele/language-formatters-pre-commit-hooks
+    rev: v2.14.0
+    hooks:
+    -   id: pretty-format-toml
+        args: [--autofix]

CHANGELOG.md

@@ -6,21 +6,33 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [0.18.0] - MM/DD/2025
 
+This release achieves 100% compliance with Python Array API specification (revision [2024.12](https://data-apis.org/array-api/2024.12/)).
+
 ### Added
 
 * Added implementation of `dpnp.hamming` [#2341](https://github.com/IntelPython/dpnp/pull/2341), [#2357](https://github.com/IntelPython/dpnp/pull/2357)
 * Added implementation of `dpnp.hanning` [#2358](https://github.com/IntelPython/dpnp/pull/2358)
 * Added implementation of `dpnp.blackman` [#2363](https://github.com/IntelPython/dpnp/pull/2363)
 * Added implementation of `dpnp.bartlett` [#2366](https://github.com/IntelPython/dpnp/pull/2366)
+* Added implementation of `dpnp.convolve` [#2205](https://github.com/IntelPython/dpnp/pull/2205)
+* Added implementation of `dpnp.kaiser` [#2387](https://github.com/IntelPython/dpnp/pull/2387)
 
 ### Changed
 
+* Improved performance of `dpnp.nansum`, `dpnp.nanprod`, `dpnp.nancumsum`, and `dpnp.nancumprod` by reusing `dpnp.nan_to_num` function in implementation of the functions [#2339](https://github.com/IntelPython/dpnp/pull/2339)
 * Allowed input array of `uint64` dtype in `dpnp.bincount` [#2361](https://github.com/IntelPython/dpnp/pull/2361)
 * The vector norms `ord={None, 1, 2, inf}` and the matrix norms `ord={None, 1, 2, inf, "fro", "nuc"}` now consistently return zero for empty arrays, which are arrays with at least one axis of size zero. This change affects `dpnp.linalg.norm`, `dpnp.linalg.vector_norm`, and `dpnp.linalg.matrix_norm`. Previously, dpnp would either raise errors or return zero depending on the parameters provided [#2371](https://github.com/IntelPython/dpnp/pull/2371)
+* Extended `dpnp.fft.fftfreq` and `dpnp.fft.rfftfreq` functions to support `dtype` keyword per Python Array API spec 2024.12 [#2384](https://github.com/IntelPython/dpnp/pull/2384)
+* Updated `dpnp.fix` to return output with the same data-type of input [#2392](https://github.com/IntelPython/dpnp/pull/2392)
+* Updated `dpnp.einsum` to add support for `order=None` [#2411](https://github.com/IntelPython/dpnp/pull/2411)
+* Updated Python Array API specification version supported to `2024.12` [#2416](https://github.com/IntelPython/dpnp/pull/2416)
+* Removed `einsum_call` keyword from `dpnp.einsum_path` signature [#2421](https://github.com/IntelPython/dpnp/pull/2421)
 
 ### Fixed
 
 * Resolved an issue with an incorrect result returned due to missing dependency from the strided kernel on a copy event in `dpnp.erf` [#2378](https://github.com/IntelPython/dpnp/pull/2378)
+* Updated `conda create` commands build and install instructions of `Quick start guide` to avoid a compilation error [#2395](https://github.com/IntelPython/dpnp/pull/2395)
+* Added handling of empty string passed to a test env variable defining data type scope as a `False` value [#2415](https://github.com/IntelPython/dpnp/pull/2415)
 
 
 ## [0.17.0] - 02/26/2025

conda-recipe/meta.yaml

@@ -1,6 +1,6 @@
 {% set max_compiler_and_mkl_version = environ.get("MAX_BUILD_CMPL_MKL_VERSION", "2026.0a0") %}
 {% set required_compiler_and_mkl_version = "2025.0" %}
-{% set required_dpctl_version = "0.19.0*" %}
+{% set required_dpctl_version = "0.20.0*" %}
 
 package:
     name: dpnp

doc/quick_start_guide.rst

@@ -88,14 +88,18 @@ On Linux:
 
 .. code-block:: bash
 
-    conda create -n build-env dpctl cython dpcpp_linux-64 mkl-devel-dpcpp tbb-devel onedpl-devel cmake scikit-build ninja pytest -c https://software.repos.intel.com/python/conda/ -c conda-forge
+    conda create -n build-env dpctl cython dpcpp_linux-64 mkl-devel-dpcpp tbb-devel \
+          onedpl-devel cmake scikit-build ninja pytest intel-gpu-ocl-icd-system     \
+          -c dppy/label/dev -c https://software.repos.intel.com/python/conda/ -c conda-forge
     conda activate build-env
 
 On Windows:
 
 .. code-block:: bash
 
-    conda create -n build-env dpctl cython dpcpp_win-64 mkl-devel-dpcpp tbb-devel onedpl-devel cmake scikit-build ninja pytest -c https://software.repos.intel.com/python/conda/ -c conda-forge
+    conda create -n build-env dpctl cython dpcpp_win-64 mkl-devel-dpcpp tbb-devel \
+          onedpl-devel cmake scikit-build ninja pytest intel-gpu-ocl-icd-system   \
+          -c dppy/label/dev -c https://software.repos.intel.com/python/conda/ -c conda-forge
     conda activate build-env
 
 To build and install the package on Linux OS, run:

doc/reference/array_api.rst

@@ -8,7 +8,7 @@ Array API standard compatibility
 
 DPNP's main namespace as well as the :mod:`dpnp.fft` and :mod:`dpnp.linalg`
 namespaces are compatible with the
-`2023.12 version <https://data-apis.org/array-api/2023.12/index.html>`__
+`2024.12 version <https://data-apis.org/array-api/2024.12/index.html>`__
 of the Python array API standard.
 
 Inspection

dpnp/backend/extensions/blas/blas_py.cpp

@@ -62,8 +62,7 @@ PYBIND11_MODULE(_blas_impl, m)
     using event_vecT = std::vector<sycl::event>;
 
     {
-        dot_ns::init_dot_dispatch_vector<dot_impl_fn_ptr_t,
-                                         blas_ns::DotContigFactory>(
+        dot_ns::init_dot_dispatch_vector<blas_ns::DotContigFactory>(
             dot_dispatch_vector);
 
         auto dot_pyapi = [&](sycl::queue &exec_q, const arrayT &src1,
@@ -81,8 +80,7 @@ PYBIND11_MODULE(_blas_impl, m)
     }
 
     {
-        dot_ns::init_dot_dispatch_vector<dot_impl_fn_ptr_t,
-                                         blas_ns::DotcContigFactory>(
+        dot_ns::init_dot_dispatch_vector<blas_ns::DotcContigFactory>(
             dotc_dispatch_vector);
 
         auto dotc_pyapi = [&](sycl::queue &exec_q, const arrayT &src1,
@@ -101,8 +99,7 @@ PYBIND11_MODULE(_blas_impl, m)
     }
 
     {
-        dot_ns::init_dot_dispatch_vector<dot_impl_fn_ptr_t,
-                                         blas_ns::DotuContigFactory>(
+        dot_ns::init_dot_dispatch_vector<blas_ns::DotuContigFactory>(
             dotu_dispatch_vector);
 
         auto dotu_pyapi = [&](sycl::queue &exec_q, const arrayT &src1,

dpnp/backend/extensions/blas/dot_common.hpp

@@ -50,14 +50,13 @@ typedef sycl::event (*dot_impl_fn_ptr_t)(sycl::queue &,
 namespace dpctl_td_ns = dpctl::tensor::type_dispatch;
 namespace py = pybind11;
 
-template <typename dispatchT>
 std::pair<sycl::event, sycl::event>
     dot_func(sycl::queue &exec_q,
              const dpctl::tensor::usm_ndarray &vectorX,
              const dpctl::tensor::usm_ndarray &vectorY,
              const dpctl::tensor::usm_ndarray &result,
              const std::vector<sycl::event> &depends,
-             const dispatchT &dot_dispatch_vector)
+             const dot_impl_fn_ptr_t *dot_dispatch_vector)
 {
     const int vectorX_nd = vectorX.get_ndim();
     const int vectorY_nd = vectorY.get_ndim();
@@ -166,12 +165,10 @@ std::pair<sycl::event, sycl::event>
     return std::make_pair(args_ev, dot_ev);
 }
 
-template <typename dispatchT,
-          template <typename fnT, typename T>
-          typename factoryT>
-void init_dot_dispatch_vector(dispatchT dot_dispatch_vector[])
+template <template <typename fnT, typename T> typename factoryT>
+void init_dot_dispatch_vector(dot_impl_fn_ptr_t dot_dispatch_vector[])
 {
-    dpctl_td_ns::DispatchVectorBuilder<dispatchT, factoryT,
+    dpctl_td_ns::DispatchVectorBuilder<dot_impl_fn_ptr_t, factoryT,
                                        dpctl_td_ns::num_types>
         contig;
     contig.populate_dispatch_vector(dot_dispatch_vector);

dpnp/backend/extensions/statistics/common.hpp renamed to dpnp/backend/extensions/common/ext/common.hpp

@@ -35,7 +35,7 @@
 
 namespace type_utils = dpctl::tensor::type_utils;
 
-namespace statistics::common
+namespace ext::common
 {
 
 template <typename N, typename D>
@@ -185,4 +185,6 @@ sycl::nd_range<1>
 // headers of dpctl.
 pybind11::dtype dtype_from_typenum(int dst_typenum);
 
-} // namespace statistics::common
+} // namespace ext::common
+
+#include "ext/details/common_internal.hpp"

dpnp/backend/extensions/statistics/common.cpp renamed to dpnp/backend/extensions/common/ext/details/common_internal.hpp

@@ -23,15 +23,15 @@
 // THE POSSIBILITY OF SUCH DAMAGE.
 //*****************************************************************************
 
-#include "common.hpp"
+#include "ext/common.hpp"
 #include "utils/type_dispatch.hpp"
 #include <pybind11/pybind11.h>
 
 namespace dpctl_td_ns = dpctl::tensor::type_dispatch;
 
-namespace statistics::common
+namespace ext::common
 {
-size_t get_max_local_size(const sycl::device &device)
+inline size_t get_max_local_size(const sycl::device &device)
 {
     constexpr const int default_max_cpu_local_size = 256;
     constexpr const int default_max_gpu_local_size = 0;
@@ -40,9 +40,9 @@ size_t get_max_local_size(const sycl::device &device)
                               default_max_gpu_local_size);
 }
 
-size_t get_max_local_size(const sycl::device &device,
-                          int cpu_local_size_limit,
-                          int gpu_local_size_limit)
+inline size_t get_max_local_size(const sycl::device &device,
+                                 int cpu_local_size_limit,
+                                 int gpu_local_size_limit)
 {
     int max_work_group_size =
         device.get_info<sycl::info::device::max_work_group_size>();
@@ -56,30 +56,31 @@ size_t get_max_local_size(const sycl::device &device,
     return max_work_group_size;
 }
 
-sycl::nd_range<1>
+inline sycl::nd_range<1>
     make_ndrange(size_t global_size, size_t local_range, size_t work_per_item)
 {
     return make_ndrange(sycl::range<1>(global_size),
                         sycl::range<1>(local_range),
                         sycl::range<1>(work_per_item));
 }
 
-size_t get_local_mem_size_in_bytes(const sycl::device &device)
+inline size_t get_local_mem_size_in_bytes(const sycl::device &device)
 {
     // Reserving 1kb for runtime needs
     constexpr const size_t reserve = 1024;
 
     return get_local_mem_size_in_bytes(device, reserve);
 }
 
-size_t get_local_mem_size_in_bytes(const sycl::device &device, size_t reserve)
+inline size_t get_local_mem_size_in_bytes(const sycl::device &device,
+                                          size_t reserve)
 {
     size_t local_mem_size =
         device.get_info<sycl::info::device::local_mem_size>();
     return local_mem_size - reserve;
 }
 
-pybind11::dtype dtype_from_typenum(int dst_typenum)
+inline pybind11::dtype dtype_from_typenum(int dst_typenum)
 {
     dpctl_td_ns::typenum_t dst_typenum_t =
         static_cast<dpctl_td_ns::typenum_t>(dst_typenum);
@@ -117,4 +118,4 @@ pybind11::dtype dtype_from_typenum(int dst_typenum)
     }
 }
 
-} // namespace statistics::common
+} // namespace ext::common