Upgrade minimum python version from 3.8 to 3.10

.github/workflows/basic_test.yml

@@ -11,7 +11,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: ["3.8"]
+        python-version: ["3.10"]
 
     steps:
       - uses: actions/checkout@v3

.github/workflows/python-publish.yml

@@ -22,7 +22,7 @@ jobs:
     - name: Set up Python
       uses: actions/setup-python@v2
       with:
-        python-version: '3.8'
+        python-version: '3.10'
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip

README.md

@@ -9,7 +9,7 @@ LibMultiLabel is a library for binary, multi-class, and multi-label classificati
 This is an on-going development so many improvements are still being made. Comments are very welcome.
 
 ## Environments
-- Python: 3.8+
+- Python: 3.10+
 - CUDA: 11.8, 12.1 (if training neural networks by GPU)
 - Pytorch: 2.0.1+
 

docs/cli/nn.rst

@@ -133,7 +133,8 @@ To deploy/evaluate a model (i.e., a pre-obtained checkpoint), you can predict a
 Hyper-parameter Search
 ^^^^^^^^^^^^^^^^^^^^^^
 
-Parameter selection is known to be extremely important in machine learning practice; see a powerful reminder in "`this paper <https://www.csie.ntu.edu.tw/~cjlin/papers/parameter_selection/acl2021_parameter_selection.pdf>`_". Here we leverage `Ray Tune <https://docs.ray.io/en/master/tune/index.html>`__, which is a python library for hyper-parameter tuning, to select parameters. Due to the dependency of Ray Tune, first make sure your python version is not greater than 3.8. Then, install the related packages with::
+Parameter selection is known to be extremely important in machine learning practice; see a powerful reminder in "`this paper <https://www.csie.ntu.edu.tw/~cjlin/papers/parameter_selection/acl2021_parameter_selection.pdf>`_". 
+Here we leverage `Ray Tune <https://docs.ray.io/en/master/tune/index.html>`__, which is a python library for hyper-parameter tuning, to select parameters. Install the related packages with::
 
     pip3 install -Ur requirements_parameter_search.txt
 

docs/cli/ov_data_format.rst

@@ -24,7 +24,7 @@ Install LibMultiLabel from Source
 
 * Environment
 
-    * Python: 3.8+
+    * Python: 3.10+
     * CUDA: 11.8, 12.1 (if training neural networks by GPU)
     * Pytorch 2.0.1+
 
@@ -37,7 +37,7 @@ and then create a virtual enviroment as follows.
 
 .. code-block:: bash
 
-    conda create -n LibMultiLabel python=3.8
+    conda create -n LibMultiLabel python=3.10
     conda activate LibMultiLabel
 
 * Clone `LibMultiLabel <https://github.com/ntumlgroup/LibMultiLabel>`_.

docs/readme

@@ -1,4 +1,4 @@
-To build the documentation, please set up Python 3.8 and:
+To build the documentation, please set up Python 3.10 and:
     pip install -r ../requirements.txt
     pip install -r ../requirements_nn.txt
     pip install -r requirements.txt

libmultilabel/linear/linear.py

@@ -27,7 +27,7 @@ class FlatModel:
     def __init__(
         self,
         name: str,
-        weights: np.matrix,
+        weights: np.matrix | sparse.csr_matrix,
         bias: float,
         thresholds: float | np.ndarray,
         multiclass: bool,
@@ -69,7 +69,21 @@ def predict_values(self, x: sparse.csr_matrix) -> np.ndarray:
                 "csr",
             )
 
-        return (x * self.weights).A + self.thresholds
+        return self._to_dense_array(x * self.weights) + self.thresholds
+
+    def _to_dense_array(self, matrix: np.matrix | sparse.csr_matrix) -> np.ndarray:
+        """Convert a numpy or scipy matrix to a dense ndarray.
+
+        Args:
+            matrix (np.matrix | sparse.csr_matrix): A numpy or scipy sparse matrix.
+
+        Returns:
+            np.ndarray: A dense ndarray of `matrix`.
+        """
+        if sparse.issparse(matrix):
+            return matrix.toarray()
+        elif isinstance(matrix, np.matrix):
+            return matrix.A
 
 
 def train_1vsrest(
@@ -458,7 +472,7 @@ def _cost_sensitive_one_label(y: np.ndarray, x: sparse.csr_matrix, options: str)
 
     param_space = [1, 1.33, 1.8, 2.5, 3.67, 6, 13]
 
-    bestScore = -np.Inf
+    bestScore = -np.inf
     for a in param_space:
         cv_options = f"{options} -w1 {a}"
         pred = _cross_validate(y, x, cv_options, perm)
@@ -532,7 +546,7 @@ def train_cost_sensitive_micro(
     l = y.shape[0]
     perm = np.random.permutation(l)
     param_space = [1, 1.33, 1.8, 2.5, 3.67, 6, 13]
-    bestScore = -np.Inf
+    bestScore = -np.inf
 
     if verbose:
         logging.info(f"Training cost-sensitive model for Micro-F1 on {num_class} labels")

requirements.txt

@@ -3,6 +3,6 @@ numba
 pandas>1.3.0
 PyYAML
 scikit-learn
-scipy<1.14.0
+scipy
 tqdm
 psutil

requirements_nn.txt

@@ -1,8 +1,8 @@
 nltk
-# wait for https://github.com/Lightning-AI/pytorch-lightning/pull/19191
-lightning==2.0.9
+lightning
 # https://github.com/pytorch/text/releases
 torch<=2.3
 torchmetrics==0.10.3
 torchtext
-transformers
+# https://github.com/huggingface/transformers/issues/38464
+transformers<=4.51.3

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = libmultilabel
-version = 0.7.4
+version = 0.8.0
 author = LibMultiLabel Team
 license = MIT License
 license_file = LICENSE
@@ -20,7 +20,7 @@ classifiers =
     License :: OSI Approved :: MIT License
     Operating System :: OS Independent
     Programming Language :: Python :: 3
-    Programming Language :: Python :: 3.8
+    Programming Language :: Python :: 3.10
 
 [options]
 packages = find:
@@ -30,20 +30,20 @@ install_requires =
     pandas>1.3.0
     PyYAML
     scikit-learn
-    scipy<1.14.0
+    scipy
     tqdm
     psutil
 
-python_requires = >=3.8
+python_requires = >=3.10
 
 [options.extras_require]
 nn =
-    lightning==2.0.9
+    lightning
     nltk
     torch<=2.3
     torchmetrics==0.10.3
     torchtext
-    transformers
+    transformers<=4.51.3
 
 
 [options.packages.find]