feat: proposed clustering method and enhance visualizations.

- Updated the pre-commit configuration to enforce stricter type checking with MyPy by adding the '--ignore-missing-imports' and '--strict' flags.
- Modified the pyproject.toml to include new dependencies for HDBSCAN and scikit-learn, enhancing clustering functionality.
- Introduced a new dataset 'blobs_300_4_3_1.0.csv' for clustering experiments, providing a synthetic dataset for testing.
- Added a new Jupyter notebook 'clustering.ipynb' for clustering analysis using Self-Organizing Maps (SOM), demonstrating various clustering methods and visualizations.
- Implemented clustering algorithms in 'torchsom/utils/clustering.py' and integrated them into the SOM class, allowing for flexible clustering options.
- Developed a 'ClusteringVisualizer' class in 'torchsom/visualization/clustering.py' to facilitate the visualization of clustering results, including elbow analysis and silhouette scores.
- Updated existing visualization methods in the SOMVisualizer to incorporate clustering visualizations, improving the overall user experience.

Author: LouisTier

.pre-commit-config.yaml

@@ -44,9 +44,8 @@ repos:
     hooks:
       - id: mypy
         name: mypy
-        args: ['--config-file', 'pyproject.toml']
+        args: ['--config-file', 'pyproject.toml', '--ignore-missing-imports', '--strict']
         # additional_dependencies: [torch, numpy, pydantic, types-requests]
-        # args: [--ignore-missing-imports, --strict]
 
   - repo: https://github.com/PyCQA/bandit
     rev: 1.7.5