Official Support for Deep Forest with modern Python versions 3.10+

.github/workflows/build-and-test.yml

@@ -12,11 +12,11 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest, windows-latest, macos-latest]
-        python-version: [3.7, 3.8, 3.9]
+        python-version: ["3.10", "3.11", "3.12"]
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v4
     - name: Set up Python
-      uses: actions/setup-python@v2
+      uses: actions/setup-python@v5
       with:
         python-version: ${{ matrix.python-version }}
     - name: Display python version
@@ -33,8 +33,9 @@ jobs:
     - name: Run tests
       run: |
         pytest ./tests --cov-config=.coveragerc --cov-report=xml --cov=deepforest deepforest
-    - name: Publish code coverage
-      uses: codecov/codecov-action@v1
-      with:
-        token: ${{ secrets.CODECOV_TOKEN }}
-        file: ./coverage.xml
+    # This step requires CODECOV_TOKEN, which is only avail. in the original repo. We therefore comment that out.
+    # - name: Publish code coverage
+    #   uses: codecov/codecov-action@v5
+    #   with:
+    #     token: ${{ secrets.CODECOV_TOKEN }}
+    #     file: ./coverage.xml

.github/workflows/build-wheels.yml

@@ -14,25 +14,28 @@ jobs:
       matrix:
         os: [ubuntu-latest, windows-latest, macos-latest]
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v4
 
       - name: Set up QEMU
         if: runner.os == 'Linux'
-        uses: docker/setup-qemu-action@v1
+        uses: docker/setup-qemu-action@v3
         with:
           platforms: all
 
       - name: Build wheels
-        uses: joerick/cibuildwheel@v1.9.0
+        uses: pypa/cibuildwheel@v2.16.5
         with:
           output-dir: wheelhouse
         env:
           CIBW_ARCHS_LINUX: "x86_64 aarch64"
           CIBW_ARCHS_WINDOWS: "AMD64"
           CIBW_ARCHS_MACOS: "x86_64"
-          CIBW_BUILD: cp3*-macosx_x86_64 cp3*-win_amd64 cp3*-manylinux_x86_64 cp3*-manylinux_aarch64
-          CIBW_SKIP: cp35-* cp36-*
+          CIBW_BUILD: >-
+            cp310-manylinux_x86_64 cp310-manylinux_aarch64 cp310-macosx_x86_64 cp310-win_amd64
+            cp311-manylinux_x86_64 cp311-manylinux_aarch64 cp311-macosx_x86_64 cp311-win_amd64
+            cp312-manylinux_x86_64 cp312-manylinux_aarch64 cp312-macosx_x86_64 cp312-win_amd64
       - name: Store artifacts
-        uses: actions/upload-artifact@v2
+        uses: actions/upload-artifact@v4
         with:
+          name: wheels-${{ matrix.os }}
           path: ./wheelhouse/*.whl

.github/workflows/code-quality.yml

@@ -12,11 +12,11 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest]
-        python-version: [3.7]
+        python-version: ["3.10"]
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v4
     - name: Set up python
-      uses: actions/setup-python@v2
+      uses: actions/setup-python@v5
       with:
         python-version: ${{ matrix.python-version }}
     - name: Display python version

.github/workflows/manual-release.yml

@@ -0,0 +1,84 @@
+name: Manual Release
+
+on:
+  workflow_dispatch:
+    inputs:
+      version:
+        description: "Version to release (e.g., 0.1.8 <~ Must be above 0.1.7 as this is the point where we forked DF21)"
+        required: true
+
+permissions:
+  contents: write
+
+jobs:
+  release:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+
+      - name: Install build backend
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install build toml
+
+      - name: Update version in setup.py
+        env:
+          RELEASE_VERSION: "${{ github.event.inputs.version }}"
+        run: |
+          python - <<'PY'
+          import os
+          import pathlib
+          import re
+
+          release_version = os.environ["RELEASE_VERSION"].strip()
+          if not release_version:
+              raise SystemExit("RELEASE_VERSION is required")
+
+          version = pathlib.Path("setup.py")
+          text = version.read_text()
+          new_text, count = re.subn(
+              r'^VERSION = "[^"]+"', f'VERSION = "{release_version}"', text, flags=re.MULTILINE
+          )
+          if count == 0:
+              raise SystemExit("VERSION assignment not found in setup.py")
+          if new_text != text:
+              version.write_text(new_text)
+          PY
+        shell: bash
+
+      - name: Commit version bump
+        env:
+          RELEASE_VERSION: "${{ github.event.inputs.version }}"
+        run: |
+          git config user.name "github-actions[bot]"
+          git config user.email "github-actions[bot]@users.noreply.github.com"
+          git add setup.py
+          if git diff --cached --quiet; then
+            echo "No changes to commit"
+          else
+            git commit -m "chore: release v${RELEASE_VERSION}"
+          fi
+
+      - name: Create and push tag
+        env:
+          RELEASE_VERSION: "${{ github.event.inputs.version }}"
+        run: |
+          git tag "v${RELEASE_VERSION}" || true
+          git push origin HEAD
+          git push origin "v${RELEASE_VERSION}"
+
+      - name: Build distributions
+        run: python -m build
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          password: ${{ secrets.PYPI_API_TOKEN }}

README.rst

@@ -1,6 +1,10 @@
 Deep Forest (DF) 21
 ===================
 
+.. note::
+
+   This repository is a community-maintained copy of the original `Deep Forest (DF21) project <https://github.com/LAMDA-NJU/Deep-Forest>`_. We do **not** claim credit for the underlying research or implementation; our goal is to keep the project usable on modern Python versions (e.g., Python 3.10+) while the upstream repository has seen minimal activity since a long time (with one recent try to move to py310+ but with CI/CD failing). If the upstream maintainers prioritise these updates, we are happy to contribute everything back via pull request.
+
 |github|_ |readthedocs|_ |codecov|_ |python|_ |pypi|_ |style|_
 
 .. |github| image:: https://github.com/LAMDA-NJU/Deep-Forest/workflows/DeepForest-CI/badge.svg

deepforest/cascade.py

@@ -350,10 +350,10 @@ def _build_regressor_predictor(
         The maximum number of cascade layers in the deep forest. Notice that
         the actual number of layers can be smaller than ``max_layers`` because
         of the internal early stopping stage.
-    criterion : :obj:`{"mse", "mae"}`, default= :obj:`"mse"`
-        The function to measure the quality of a split. Supported criteria are 
-        ``mse`` for the mean squared error, which is equal to variance reduction 
-        as feature selection criterion, and ``mae`` for the mean absolute error.
+    criterion : :obj:`{"squared_error", "absolute_error"}`, default= :obj:`"squared_error"`
+        The function to measure the quality of a split. Supported criteria are
+        ``squared_error`` for the mean squared error, which is equal to variance reduction
+        as feature selection criterion, and ``absolute_error`` for the mean absolute error.
     n_estimators : :obj:`int`, default=2
         The number of estimator in each cascade layer. It will be multiplied
         by 2 internally because each estimator contains a
@@ -1553,7 +1553,7 @@ def __init__(
         bin_subsample=200000,
         bin_type="percentile",
         max_layers=20,
-        criterion="mse",
+        criterion="squared_error",
         n_estimators=2,
         n_trees=100,
         max_depth=None,

deepforest/forest.py

@@ -49,6 +49,17 @@
 MAX_INT = np.iinfo(np.int32).max
 
 
+def _normalize_regression_criterion(criterion: str) -> str:
+    """Align legacy regression criteria names with scikit-learn
+    expectations."""
+
+    if criterion == "mse":
+        return "squared_error"
+    if criterion == "mae":
+        return "absolute_error"
+    return criterion
+
+
 def _get_n_samples_bootstrap(n_samples, max_samples):
     """
     Get the number of samples in a bootstrap sample.
@@ -199,9 +210,7 @@ def _partition_estimators(n_estimators, n_jobs):
     n_jobs = min(effective_n_jobs(n_jobs), n_estimators)
 
     # Partition estimators between jobs
-    n_estimators_per_job = np.full(
-        n_jobs, n_estimators // n_jobs, dtype=int
-    )
+    n_estimators_per_job = np.full(n_jobs, n_estimators // n_jobs, dtype=int)
     n_estimators_per_job[: n_estimators % n_jobs] += 1
     starts = np.cumsum(n_estimators_per_job)
 
@@ -840,7 +849,7 @@ def __init__(
         self,
         n_estimators=100,
         *,
-        criterion="mse",
+        criterion="squared_error",
         max_depth=None,
         min_samples_split=2,
         min_samples_leaf=1,
@@ -853,6 +862,7 @@ def __init__(
         verbose=0,
         max_samples=None
     ):
+        criterion = _normalize_regression_criterion(criterion)
         super().__init__(
             base_estimator=DecisionTreeRegressor(),
             n_estimators=n_estimators,
@@ -889,7 +899,7 @@ def __init__(
         self,
         n_estimators=100,
         *,
-        criterion="mse",
+        criterion="squared_error",
         max_depth=None,
         min_samples_split=2,
         min_samples_leaf=1,
@@ -902,6 +912,7 @@ def __init__(
         verbose=0,
         max_samples=None
     ):
+        criterion = _normalize_regression_criterion(criterion)
         super().__init__(
             base_estimator=ExtraTreeRegressor(),
             n_estimators=n_estimators,

deepforest/tree/tree.py

@@ -49,7 +49,12 @@
 DOUBLE = _tree.DOUBLE
 
 CRITERIA_CLF = {"gini": _criterion.Gini, "entropy": _criterion.Entropy}
-CRITERIA_REG = {"mse": _criterion.MSE, "mae": _criterion.MAE}
+CRITERIA_REG = {
+    "mse": _criterion.MSE,
+    "mae": _criterion.MAE,
+    "squared_error": _criterion.MSE,
+    "absolute_error": _criterion.MAE,
+}
 
 DENSE_SPLITTERS = {
     "best": _splitter.BestSplitter,
@@ -181,7 +186,7 @@ def fit(
             if self.class_weight is not None:
                 y_original = np.copy(y)
 
-            y_encoded = np.zeros(y.shape, dtype=np.int)
+            y_encoded = np.zeros(y.shape, dtype=int)
             for k in range(self.n_outputs_):
                 classes_k, y_encoded[:, k] = np.unique(
                     y[:, k], return_inverse=True
@@ -504,7 +509,7 @@ class DecisionTreeRegressor(RegressorMixin, BaseDecisionTree):
     def __init__(
         self,
         *,
-        criterion="mse",
+        criterion="squared_error",
         splitter="best",
         max_depth=None,
         min_samples_split=2,
@@ -580,7 +585,7 @@ class ExtraTreeRegressor(DecisionTreeRegressor):
     def __init__(
         self,
         *,
-        criterion="mse",
+        criterion="squared_error",
         splitter="random",
         max_depth=None,
         min_samples_split=2,

pyproject.toml

@@ -1,11 +1,13 @@
 [build-system]
 requires = [
-    "setuptools<60.0",
+    "setuptools>=61,<70",
     "wheel",
+    "numpy==1.26.4",
     "Cython>=0.28.5,<3.0",
     "oldest-supported-numpy",
     "scipy>=1.3.2",
 ]
+build-backend = "setuptools.build_meta"
 [tool.black]
 line-length = 79
 include = '\.pyi?$'