Releases: romerogroup/MatGraphDB
Releases · romerogroup/MatGraphDB
v0.3.0
0.3.0 (02-08-2025)
Bugs
- None identified
New Features
- Implemented heterogeneous graph encoder training infrastructure
- Added PyTorch Geometric (PyG) infrastructure for graph machine learning
- Included new periodic table values and coordination geometries to support advanced material analysis
- Added new dataset initialization in the materials.datasets module to support the new MPNearHull functionality
Documentation
- Updated _version.py and CHANGELOG.md due to new release
Maintenance
- Streamlined dependency management in pyproject.toml
- Improved initialization logic to handle dataset presence and downloading options
- Updated .gitignore to exclude the /data/external directory, streamlining data management and ensuring that only relevant data directories are tracked
- Added .gitattributes to track data
- Added comprehensive tests for the new GraphBuilder functionality and updated existing tests to reflect changes in edge data structure
v0.2.0
0.2.0 (01-13-2025)
Bugs
- None identified
New features
- None identified
Documentation updates
- None identified
Maintenance
- Improved logging and error handling across various modules for better traceability during operations
- Updated _version.py and CHANGELOG.md due to new release
0.1.0
0.1.0 (01-11-2025)
Bugs
- Fixed bugs in MatDB and add tests for MatDB
- Changes due to the API change in MatDB and changes to ParquetDB
- bug: typo in env.tml
New Features
- Added CalculationManager and core utilities for material calculations
- Add MatGraphDB class for advanced material analysis and management
- Add MaterialStore class for comprehensive material management
- Add custom PyArrow classes for enhanced material structure handling
- Added method to get element_properties
- Added better docs to coord_geometry module
- Added sc 8000 ript to handle loading of coordination data previously stored in JSON, now in Parquet format
Documentation
- Update README.md to streamline the introduction and improve clarity
- Refactor Nodes class documentation in nodes.py
- Updated to have matgraphdb to have version variable
- Update _version.py and CHANGELOG.md due to new release
Maintenance
- Remove deprecated files and scripts from the sandbox and old_scripts directories
- Improved logging and error handling across various modules
- Refactor edge and node management in MatGraphDB
- Refactor
NodeStoreusage across multiple modules - Removed old test data
- Cleaned up whitespace in the SpaceGroupNodes class
- Refactor GraphStore methods for improved clarity and consistency
- Refactor GraphStore initialization and enhance directory structure
- Refactor MaterialStore for improved data management and functionality
- Refactor EdgeStore and GraphStore initialization logic
- Update .gitignore, removed examples
- Moved old sandbox directories to old_scripts
- Removed old chem util resource formats
- Updated environmental configuration management
- Updated directory structure for examples
- Updated publish workflow
- Updated dependencies
- Updated GitHub workflow scripts
- Enhanced configuration and logging settings in MatGraphDB
v0.0.3
MatGraphDB v0.0.3 - New Deployment Workflow and Cleanup
What's New:
This release introduces a new automated Python deployment workflow along with some cleanup and reorganization of the repository.
Key Changes:
-
New Python Deployment Workflow:
- On every release, this workflow will:
- Build and deploy the package to PyPI.
- Automatically generate a changelog based on git commits.
- Update the repository version.
- Update the release notes with the current release details.
This addition simplifies the release process and ensures consistency for future deployments.
- On every release, this workflow will:
-
File Cleanup and Organization:
- Moved Old Scripts to Sandbox: Legacy scripts have been moved to a sandbox directory for archival purposes.
- Removed Unused File: A redundant or unused file has been removed, reducing clutter in the codebase.
0.0.3 (10-03-2024)
Bugs
- None identified
New Features
- Added new python deployment workflow. On release, this will build and deploy on PyPI, generate a changelog from the commits, update the repo version, and update the release notes with the current change version changelog
Documentation updates
- None identified
Maintenance
- Moved old scripts to sandbox
- Removed unused file