v1.2.0

[1.2.0] - 2026-06-24

Added

• Added laminar flame speed support: freely-propagating FlameSimulation, the laminar-flame-conditions input block, and laminar flame speed as an error metric, enabling reductions driven by flame data alone or combined with autoignition (based on work by Cailin Moore, #82)
• Added BaseSimulation abstract base class with shared gas setup and thermochemical-profile sampling, giving IgnitionSimulation and FlameSimulation a common interface
• Added automatic retention of fuel, oxidizer, and reactant species from the initial conditions so they are never removed during reduction (fixes #12; based on work by Katherine Bronstein, #80)
• Added soln2yaml module for writing Cantera Solution objects to YAML files, delegating to Cantera's built-in write_yaml() method
• Added YAML equivalents of all test asset models (artificial-mechanism.yaml, model-third-bodies.yaml, reduced GRI 3.0 models)
• Added conda recipe (conda.recipe/meta.yaml) updated for Python 3.10+ and Cantera 3.x
• Added GitHub Actions CI workflow with matrix testing on Python 3.10–3.14 (Linux) and Python 3.14 (macOS, Windows), with Coveralls coverage reporting
• Added GitHub Actions publish workflow for PyPI (OIDC trusted publishing) and Anaconda.org; conda publish triggers automatically on v* tags after CI passes
• Added GitHub Actions docs workflow that deploys to GitHub Pages after CI passes on main

Changed

• Autoignition simulations now use Cantera mole reactors (IdealGasMoleReactor / IdealGasConstPressureMoleReactor) with an adaptive preconditioner, accelerating integration of large models (based on work by Anthony Walker, #84)
• Reduction drivers (run_drg, run_drgep, run_pfa, and sensitivity analysis) now accept laminar flame conditions and combine ignition and flame error metrics and sampled state data uniformly, regardless of simulation type
• Input files now require at least one of autoignition-conditions or laminar-flame-conditions (previously autoignition was mandatory)
• Renamed the Simulation class to IgnitionSimulation and consolidated the base, ignition, and flame simulation classes into a single simulation.py
• trim() now preserves the original model's transport model so reduced models remain usable for laminar flame simulations
• Minimum Python version raised from 3.6 to 3.10
• Updated Cantera dependency to 3.x; updated all API calls for Cantera 3.x (reaction type strings, stoichiometric coefficient property names, reactor constructor, phase/kinetics identifiers)
• Replaced .cti format throughout with Cantera's YAML format; soln2cti.py removed and replaced by soln2yaml.py
• Replaced pkg_resources with pathlib and importlib.metadata for Python 3.14 compatibility
• Switched build system from setuptools to hatchling
• Model file references in CLI, documentation, and examples updated from .cti to .yaml
• compare_models now uses np.isclose() for floating-point rate parameter comparisons and evaluates thermodynamic properties at reference temperatures instead of comparing raw polynomial coefficients
• docs/conf.py updated to use importlib.metadata for version detection; removed Travis CI references; updated Cantera intersphinx URL to 3.x docs

Fixed

• Fixed bug in soln2ck.py where the reaction high rate was being used instead of the reaction low rate
• Fixed Cantera 3.x deprecation warnings: reactor.thermo → reactor.phase, IdealGas/GasKinetics → ideal-gas/bulk, reactor clone=False
• Removed broken xfail and skip-decorated tests that referenced undefined functions