- Allow state references to be cloned into normal states (useful for FFS results).
- Add Python typing for FFS results.
- Distinguish stored states and number of configurations for FFS surfaces.
- Make FFS state references more useful.
- Fix rayon use in Python (parallel execution).
- Allow Python compilation without ui.
- Fix attachment order tracking.
- Fix tube canvas N/S boundaries.
- PyO3 0.21, with new Bound API.
- Fix python tests.
- Restructured API to allow separate, easier-to-use Systems and States.
- Refactored Python API.
- Using enum_dispatch rather than dynamic traits and boxes.
- Added Python plotting function.
- Improved Python typing.
- Fixed tube canvas implementation.
- Added more mutable parameters to kTAM.
- Python tests.
- Mismatch calculations for aTAM.
- Add several convenience functions to Python API.
- Improve documentation.
Version bump to fix some release automation problems.
The version number for Rust has been increased to give consistent version numbering for both Rust and Python.
- Combined the pyrgrow and rgrow repositories, to ease development and ensure version matching.
- Initial support for mid-simulation system parameter changes.
- RateStore multi-update optimizations.
- Significant refactoring of Python API, and more documentation.
- Calculation of mismatches and mismatch locations.
- Display improvements, with mismatches shown at higher display scales, and dynamic scaling.
- Parallel evolution improvements.
- Make UI no longer default for Python.
- Type improvements.
- Optimizations.
- thread_rng instead of SmallRng (easier code, no major performance change)
- Tiles as u32.
- Initial cffi interface.
- Improved UI.
- Improved errors.
- Improved evolve bounds specifications.
- Partially merged some Python interface items into this crate.
- Improved xgrow format support.
- Added tests, including the xgrow format.
- Included xgrow example files, and tested all of them.
- Other things I've probably forgotten.
- First released version.
- New kTAM model, supporting duples and changes to the system.
- New aTAM model, based on the new kTAM model.
- Simulation trait, allowing model-agnostic code.
- TileSet objects, allowing parsing of tileset inputs.