v0.3.1 and v0.3.2

.github/workflows/wheels.yml

@@ -25,9 +25,11 @@ jobs:
 
       # Used to host cibuildwheel
       - uses: actions/setup-python@v5
+        with:
+          python-version: "3.13"
 
       - name: Install cibuildwheel
-        run: python -m pip install cibuildwheel==2.20.0
+        run: python -m pip install cibuildwheel==2.22.0
 
       - name: Build wheels
         run: python -m cibuildwheel --output-dir wheelhouse

CMakeLists.txt

@@ -12,9 +12,24 @@ execute_process(
   COMMAND "${Python_EXECUTABLE}" -m nanobind --cmake_dir
   OUTPUT_STRIP_TRAILING_WHITESPACE OUTPUT_VARIABLE NB_DIR)
 list(APPEND CMAKE_PREFIX_PATH "${NB_DIR}")
+
 find_package(nanobind CONFIG REQUIRED)
+find_package(Threads REQUIRED)
+
+if(MSVC)
+  # Static link of MSVC rt for optimal compatibility
+  # It avoids to mess with embedded MSVC rt coming from other packages (see PyQT5)
+  set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreaded$<$<CONFIG:Debug>:Debug>")
+  # Same, enabling LTO (LTGC for MS) give less ABI compatibility, so we dissable it.
+  # see https://learn.microsoft.com/en-us/cpp/porting/binary-compat-2015-2017?view=msvc-170
+  # for more details.
+  nanobind_add_module(pgeof_ext NOMINSIZE STABLE_ABI src/pgeof_ext.cpp)
+else()
+  # we enable LTO on Linux and MacOS.
+  nanobind_add_module(pgeof_ext NOMINSIZE STABLE_ABI LTO src/pgeof_ext.cpp)
+endif()
 
-nanobind_add_module(pgeof_ext NOMINSIZE STABLE_ABI LTO src/pgeof_ext.cpp)
+target_link_libraries(pgeof_ext PRIVATE Threads::Threads)
 
 nanobind_add_stub(
    pgeof_ext_stub
@@ -25,10 +40,10 @@ nanobind_add_stub(
    DEPENDS pgeof_ext
 )
 
-# All lib are headeer only. 
-# it's faster to include like this than using exported targets 
+# All lib are header only.
+# it's faster to include like this than using exported targets
 # (i.e add_subdirectories(...))
-target_include_directories(pgeof_ext PRIVATE "include" "third_party/eigen" "third_party/nanoflann/include" "third_party/taskflow") 
+target_include_directories(pgeof_ext PRIVATE "include" "third_party/eigen" "third_party/nanoflann/include" "third_party/taskflow")
 
 install(TARGETS pgeof_ext LIBRARY DESTINATION pgeof)
-install(FILES ${CMAKE_CURRENT_BINARY_DIR}/pgeof_ext.pyi ${CMAKE_CURRENT_BINARY_DIR}/py.typed DESTINATION pgeof)
+install(FILES ${CMAKE_CURRENT_BINARY_DIR}/pgeof_ext.pyi ${CMAKE_CURRENT_BINARY_DIR}/py.typed DESTINATION pgeof)

pyproject.toml

@@ -1,30 +1,29 @@
 [build-system]
-requires = ["scikit-build-core >=0.4.3", 
-  "nanobind == 2.1.0",
-  "typing_extensions;python_version < '3.11'"
+requires = [
+    "scikit-build-core >=0.4.3",
+    "nanobind == 2.1.0",
+    "typing_extensions;python_version < '3.11'",
 ]
 build-backend = "scikit_build_core.build"
 
 [project]
 name = "pgeof"
-version = "0.3.0"
+version = "0.3.2"
 readme = "README.md"
 description = "Compute the geometric features associated with each point's neighborhood:"
-requires-python = ">=3.8,<3.13"
-license = {file = "LICENSE"}
+requires-python = ">=3.8,<3.14"
+license = { file = "LICENSE" }
 authors = [
-  {name = "Loic Landrieu", email = "loic.landrieu@enpc.fr"},
-  {name = "Damien Robert", email = "damien.robert@uzh.ch"}
+    { name = "Loic Landrieu", email = "loic.landrieu@enpc.fr" },
+    { name = "Damien Robert", email = "damien.robert@uzh.ch" },
 ]
 keywords = ["point clouds", "features", "3D", "LiDAR"]
 classifiers = [
-  "Development Status :: 3 - Alpha",
-  "Programming Language :: Python",
-  "Topic :: Scientific/Engineering"
-]
-dependencies = [
-  "numpy >= 1.7, < 2"
+    "Development Status :: 3 - Alpha",
+    "Programming Language :: Python",
+    "Topic :: Scientific/Engineering",
 ]
+dependencies = ["numpy >= 1.7"]
 
 [project.urls]
 homepage = "https://github.com/drprojects/point_geometric_features"
@@ -42,16 +41,16 @@ cmake.build-type = "Release"
 # make sdist a lot lighter by removing some useless files from third_party
 # ⚠️  be sure to keep copyrights and license file
 sdist.exclude = [
-  "third_party/eigen/bench",
-  "third_party/eigen/demos",
-  "third_party/eigen/doc",
-  "third_party/taskflow/3rd-party",
-  "third_party/taskflow/benchmarks",
-  "third_party/taskflow/docs",
-  "third_party/taskflow/doxygen",
-  "third_party/taskflow/examples",
-  "third_party/taskflow/sandbox",
-  "third_party/taskflow/unittests",
+    "third_party/eigen/bench",
+    "third_party/eigen/demos",
+    "third_party/eigen/doc",
+    "third_party/taskflow/3rd-party",
+    "third_party/taskflow/benchmarks",
+    "third_party/taskflow/docs",
+    "third_party/taskflow/doxygen",
+    "third_party/taskflow/examples",
+    "third_party/taskflow/sandbox",
+    "third_party/taskflow/unittests",
 ]
 
 [tool.ruff]
@@ -61,20 +60,20 @@ line-length = 120
 [tool.ruff.lint]
 # TODO Add D, PTH, RET, disabled for now as they collides with intial choices
 select = ["E", "W", "YTT", "NPY", "PYI", "Q", "F", "B", "I", "SIM", "RUF"]
-# TODO: for now we ignore "Line too long error (E501)" 
+# TODO: for now we ignore "Line too long error (E501)"
 # because our comments are too longs
 # code formatting will take care of the line length in code anyway
 ignore = [
-  "E501",
-  # Ignore docstring in public package and module
-  "D100",
-  "D104",
-  # Blank line before class
-  "D203",
-  # multiline summary second line
-  "D213",
-  # yoda conditions
-  "SIM300",
+    "E501",
+    # Ignore docstring in public package and module
+    "D100",
+    "D104",
+    # Blank line before class
+    "D203",
+    # multiline summary second line
+    "D213",
+    # yoda conditions
+    "SIM300",
 ]
 
 [tool.ruff.lint.isort]
@@ -86,28 +85,31 @@ legacy_tox_ini = """
 
 [gh-actions]
 python =
+    3.8: py39
     3.9: py39
     3.10: py310
     3.11: py311
     3.12: py312
+    3.13: py313
 
 [testenv]
 deps =
   pytest >= 7.4
   pytest-benchmark ~= 4.0
-  numpy >= 1.7, < 2
+  numpy >= 1.7
   scipy
-  jakteristics;platform_system=="Windows" or platform_system=="Linux" 
+  jakteristics;platform_system=="Windows" or platform_system=="Linux"
 commands = pytest --basetemp="{envtmpdir}" {posargs}
 
 [testenv:bench]
 # globs/wildcards do not work with tox
-commands = pytest -s --basetemp="{envtmpdir}" {posargs:tests/bench_knn.py tests/bench_jakteristics.py} 
+commands = pytest -s --basetemp="{envtmpdir}" {posargs:tests/bench_knn.py tests/bench_jakteristics.py}
 """
 
 [tool.cibuildwheel]
-build = "cp3{9,10,11,12}-*"
-archs = ["auto64"] # limit to 64bits builds
+build = "cp3{8,9,10,11,12,13}-*"
+archs = ["auto64"]               # limits to 64bits builds
+skip = "cp38-macosx_arm64"
 
 # Needed for full C++17 support
 [tool.cibuildwheel.macos.environment]

tests/bench_jakteristics.py

@@ -12,8 +12,7 @@
 @pytest.fixture
 def random_point_cloud():
     rng = np.random.default_rng()
-    return rng.random(0, 100, size=((1000000, 3)))
-
+    return rng.uniform(0.0, 200.0, size=(10000, 3))
 
 @pytest.mark.benchmark(group="feature-computation-jak", disable_gc=True, warmup=True)
 def test_bench_jak(benchmark, random_point_cloud):

tests/bench_knn.py

@@ -8,8 +8,7 @@
 @pytest.fixture
 def random_point_cloud():
     rng = np.random.default_rng()
-    return rng.random(0, 100, size=((1000000, 3))).astype("float32")
-
+    return rng.uniform(0.0, 200.0, size=(1000000, 3)).astype(np.float32)
 
 @pytest.mark.benchmark(group="knn", disable_gc=True, warmup=True)
 def test_knn_scipy(benchmark, random_point_cloud):

tests/helpers.py

@@ -5,7 +5,7 @@
 def random_nn(num_points, k):
     # Generate a random synthetic point cloud
     rng = np.random.default_rng()
-    xyz = rng.random(size=(num_points, 3))
+    xyz = rng.uniform(0.0, 200.0, size=(num_points, 3)).astype(np.float32)
 
     # Converting k-nearest neighbors to CSR format
     kneigh = KDTree(xyz).query(xyz, k=k, workers=-1)

tests/test_pgeof.py

@@ -8,8 +8,7 @@
 def test_knn():
     knn = 10
     rng = np.random.default_rng()
-    xyz = rng.random(size=(1000, 3))
-    xyz = xyz.astype("float32")
+    xyz = rng.uniform(0.0, 200.0, size=(1000, 3)).astype(np.float32)
     tree = KDTree(xyz)
     _, k_legacy = tree.query(xyz, k=knn, workers=-1)
     k_new, _ = pgeof.knn_search(xyz, xyz, knn)
@@ -20,8 +19,7 @@ def test_radius_search():
     knn = 10
     radius = 0.2
     rng = np.random.default_rng()
-    xyz = rng.random(size=(1000, 3))
-    xyz = xyz.astype("float32")
+    xyz = rng.random(size=(1000, 3), dtype=np.float32)
     tree = KDTree(xyz)
     _, k_legacy = tree.query(xyz, k=knn, distance_upper_bound=radius, workers=-1)
     k_legacy[k_legacy == xyz.shape[0]] = -1