v0.1.3.3

Resolving issue of importing C++ models on Spack.
It need to explicitly copy py files to be copied from cpp/_src.

v0.1.3.2

fixing bug for importing C++ models.

v0.1.3

Summary of New Release Based on Commit Messages (March 7, 2025)

The release reflects a mix of feature enhancements, documentation improvements, workflow optimizations, and bug fixes, The changes indicate a focus on improving usability, GPU support, Docker integration, and deployment processes.

---

Key Features and Enhancements

1. New Model Implementations and Examples:

   • Added and refined implementations of the MPR_sde model in CuPy and Numba, including example notebooks (mpr_sde_cupy.ipynb, jr_sde_cpp.ipynb, GHB_sde in CuPy, etc.).
   • Introduced a Jansen-Rit (JR) CuPy notebook example and enhanced its documentation.
   • Added new data files (myelin, rsfc_gradient, cortical, centres, my_features.json) for model inputs and statistical calculations.

2. Docker Support:

   • Refactored Docker image build workflows to support the develop branch and updated the Dockerfile to use an NVIDIA CUDA base image with Python installed.
   • Enhanced Docker usage instructions in README.md and index.rst, including port mapping and container run commands.

3. CI/CD and Deployment:

   • Added and refined GitHub Actions workflows for publishing to PyPI, including version bumping (e.g., to 0.1.3) and dynamic version retrieval.
   • Optimized workflows to specify Python versions (e.g., 3.10), streamline build processes, and improve C++ extension compilation.

4. Documentation Improvements:

   • Updated README.md, API documentation, and Read the Docs configuration with better formatting, custom CSS styles, and navigation links.
   • Added badges (Docker build, Binder) and contribution guidelines (CONTRIBUTING.md).
   • Enhanced clarity in model parameter documentation (e.g., MPR_sde class).

5. Testing and Code Quality:

   • Added unit tests for mpr_sde_numba and cpp.mpr.MPR_sde, along with error handling for C++ module imports.
   • Ignored DeprecationWarning in tests and improved GPU availability checks.

---

Bug Fixes and Refinements

• Fixed r_period calculation and RV recording conditions in the integrate function.
• Improved error messages and label prefixes in matrix_stat and fc_stat functions.

---

Workflow and Maintenance

• Refactored .gitignore, pyproject.toml, MANIFEST.in, and setup.py to include new datasets and streamline packaging.
• Updated .readthedocs.yaml with C++ dependencies and local package installation.
• Cleaned up unnecessary files (e.g., .bib) and commented out deprecated workflow steps.

---

Release Highlights

This release improves the project's usability for developers and researchers by enhancing GPU-accelerated model implementations (via CuPy/Numba), refining Docker-based deployment, and strengthening CI/CD pipelines for PyPI publishing. Documentation has been significantly polished for better accessibility, and new datasets expand the toolkit's capabilities. The changes reflect ongoing development from the develop branch, merged into the mainline via multiple pull requests (e.g., #3 to #18).

For detailed usage instructions, refer to the updated README.md and example notebooks. The current version (as of the latest commits) is likely 0.1.3, based on the version bump on February 17, 2025.

---

Let me know if you'd like a more detailed breakdown or specific sections expanded!

v0.1.2

Release Notes - 2025-02-03

Enhancements & Refactoring

• Improved CuPy utils module:
  • Replaced np.matlib.repmat with np.tile for vector repetition.
  • Optimized data movement between CPU and GPU.
• Refactored MPR_sde class:
  • Added delta array, avg_r, and RECORD_AVG_r variables.
  • Updated avg_r calculation.
• Updated preprocessing:
  • Set preprocess_args default value to {} in calc_features.py.
  • Fixed issues with feature label handling and NaN entries.
• Updated Makefile to use python3-config for PYTHON_INCLUDE.

New Features

• Added VEP model implementation in C++.
• Introduced Wilson-Cowan ODE and Wong-Wang SDE models.
• Implemented Kuramoto model SDE in C++.
• Added Stuart-Landau model in C++.
• Introduced catch22 feature extraction.
• Added slice_features functionality.

Build & Testing Updates

• Added C++ compilation step to tests.yml.
• Included SWIG as a dependency in pyproject.toml.
• Updated test workflows, renamed test files, and improved documentation build configurations.

Documentation & Miscellaneous

• Updated README.md, requirements.txt, .gitignore, and documentation files.
• Added LICENSE and MANIFEST.in files.
• Improved logging and minor utility modifications.

This release includes significant improvements in model implementations, performance optimizations, and expanded feature extraction capabilities. 🚀

virtual brain inference

update setup.py

first release

v0.1.0

pyproject.toml modified.