Merge pull request #137 from lucasimi/develop

Using numba instead of cython for specialized metrics

Author: lucasimi

.github/workflows/deploy.yml

@@ -6,9 +6,10 @@ on:
       - 'v[0-9]+.[0-9]+.[0-9]+'
 
 jobs:
-  build_locally:
-    name: Build locally
+  deploy-job:
+    name: Build and deploy
     runs-on: ubuntu-latest
+    environment: pypi
     steps:
       - name: Check out repository code
         uses: actions/checkout@v4
@@ -24,51 +25,6 @@ jobs:
         run: | 
           python -m build
           python -m twine check dist/*
-
-  build_wheels:
-    name: Build wheels on ${{ matrix.os }}
-    runs-on: ${{ matrix.os }}
-    strategy:
-      matrix:
-        os: [ubuntu-latest, windows-latest, macos-13, macos-14]
-    steps:
-      - uses: actions/checkout@v4
-      - name: Build wheels
-        uses: pypa/cibuildwheel@v2.19.2
-      - uses: actions/upload-artifact@v4
-        with:
-          name: cibw-wheels-${{ matrix.os }}-${{ strategy.job-index }}
-          path: ./wheelhouse/*.whl
-
-  build_sdist:
-    name: Build source distribution
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-
-      - name: Build sdist
-        run: pipx run build --sdist
-
-      - uses: actions/upload-artifact@v4
-        with:
-          name: cibw-sdist
-          path: dist/*.tar.gz
-
-  upload_pypi:
-    needs: [build_locally, build_wheels, build_sdist]
-    runs-on: ubuntu-latest
-    environment: pypi
-    permissions:
-      id-token: write
-    steps:
-      - uses: actions/download-artifact@v4
-        with:
-          pattern: cibw-*
-          path: dist
-          merge-multiple: true
-
-      - uses: pypa/gh-action-pypi-publish@release/v1
-        with:
-          repository-url: https://test.pypi.org/legacy/
-
-      - uses: pypa/gh-action-pypi-publish@release/v1
+      - name: Upload package to PyPI
+        run: | 
+          python -m twine upload -r pypi dist/* --verbose

pyproject.toml

@@ -1,10 +1,10 @@
 [build-system]
-requires = ["setuptools", "wheel", "Cython"]
+requires = ["setuptools>=42", "wheel"]
 build-backend = "setuptools.build_meta"
 
 [project]
 name = "tda-mapper"
-version = "0.7.1"
+version = "0.7.2"
 description = "A simple and efficient Python implementation of Mapper algorithm for Topological Data Analysis"
 readme = "README.md"
 authors = [{ name = "Luca Simi", email = "lucasimi90@gmail.com" }]
@@ -19,21 +19,19 @@ keywords = ["tda", "mapper", "topology", "topological data analysis"]
 dependencies = [
     "matplotlib>=3.3.4",
     "networkx>=2.5",
+    "numba>=0.54",
     "numpy>=1.20.1, <2.0.0",
     "plotly>=4.14.3"
 ]
-requires-python = ">=3.7"
+requires-python = ">=3.6"
 
 [project.optional-dependencies]
-dev = ["coverage", "pandas", "scikit-learn"]
+dev = ["coverage[toml]", "pandas", "scikit-learn"]
 
 [project.urls]
 Homepage = "https://github.com/lucasimi/tda-mapper-python"
 Documentation = "https://tda-mapper.readthedocs.io"
 Issues = "https://github.com/lucasimi/tda-mapper-python/issues"
 
-[tool.setuptools.package-data]
-"tdamapper.utils" = ["_metrics.c", "_metrics.pyx"]
-
-[tool.cibuildwheel]
-skip = "cp36-*"
+[tool.coverage.run]
+omit = ["**/_*.py"]

setup.py

@@ -0,0 +1,31 @@
+import numpy as np
+from numba import njit
+
+
+@njit(fastmath=True)
+def euclidean(x, y):
+    return np.linalg.norm(x - y)
+
+
+@njit(fastmath=True)
+def manhattan(x, y):
+    return np.linalg.norm(x - y, ord=1)
+
+
+@njit(fastmath=True)
+def chebyshev(x, y):
+    return np.linalg.norm(x - y, ord=np.inf)
+
+
+@njit(fastmath=True)
+def minkowski(p, x, y):
+    return np.linalg.norm(x - y, ord=p)
+
+
+@njit(fastmath=True)
+def cosine(x, y):
+    xy = np.dot(x, y)
+    xx = np.linalg.norm(x)
+    yy = np.linalg.norm(y)
+    similarity = xy / (xx * yy)
+    return np.sqrt(2.0 * (1.0 - similarity))

src/tdamapper/utils/_metrics.py

@@ -4,29 +4,37 @@
 import pandas as pd
 import numpy as np
 
+import numba
+
 import tdamapper.utils.metrics as metrics
 
 
+@numba.njit(fastmath=True)
 def euclidean_numpy(a, b):
     return np.sqrt(np.sum((a - b) ** 2))
 
 
+@numba.njit(fastmath=True)
 def euclidean_numpy_linalg(a, b):
     return np.linalg.norm(a - b)
 
 
+@numba.njit(fastmath=True)
 def manhattan_numpy(a, b):
     return np.sum(np.abs(a - b))
 
 
+@numba.njit(fastmath=True)
 def manhattan_numpy_linalg(a, b):
     return np.linalg.norm(a - b, ord=1)
 
 
+@numba.njit(fastmath=True)
 def chebyshev_numpy(a, b):
     return np.max(np.abs(a - b))
 
 
+@numba.njit(fastmath=True)
 def chebyshev_numpy_linalg(a, b):
     return np.linalg.norm(a - b, ord=np.inf)
 
@@ -37,7 +45,8 @@ def eval_dist(X, d):
 
 
 def run_dist_bench(X, d):
-    return timeit.timeit(lambda: eval_dist(X, d), number=30)
+    eval_dist(X, d)
+    return timeit.timeit(lambda: eval_dist(X, d), number=100)
 
 
 def run_euclidean_bench(X):