Version 0.0.4

Key changes

• Introduced the shared XY_Oscillator base class and ArrayParam wrapper to unify initialization, transforms, and unit-safe parameter handling.
• Added new phenomenological models (FitzHughNagumoModel, KuramotoNetwork, VanDerPolOscillator, StuartLandauOscillator, ThresholdLinearModel, QIF) plus accompanying tests for the new models, Hopf dynamics, and parameter utilities.
• Published a refreshed example suite (Hopf, Wilson–Cowan, FitzHugh–Nagumo, parameter exploration, Nevergrad/SciPy optimization, MEG modeling) with notebook and script variants, including a KaggleHub-backed MEG workflow that downloads HCP sample data on demand.
• Refactored existing dynamics (Hopf, Wilson–Cowan, Jansen–Rit, Wong–Wang, noise processes, forward model) to rely on braintools initialization utilities, expand documentation, and tighten unit-aware behavior.
• Expanded project documentation with a reorganized API reference, refreshed landing page, new project logo, and an updated README covering installation badges, ecosystem packages, and citation details.
• Simplified CI by focusing on Python 3.13 runners and removing the multi-version JAX matrix from the daily and main workflows.
• Removed bundled example datasets and legacy notebooks/scripts; the MEG example now fetches data lazily to keep the repository lightweight.

What's Changed

• Migrate HCP dataset to KaggleHub by @chaoming0625 in #10
• Add ArrayParam; Refactor parameter handling by @chaoming0625 in #11
• Add FHN, Kuramoto, VdP, Linear, QIF, SL models; Refactor Hopf and Wilson-Cowan by @chaoming0625 in #12
• Add new examples and tests for models; Refine FHN, Linear, QIF by @chaoming0625 in #13
• update logo by @xinzhu-L in #14
• Migrate initialization and optimization to braintools; Add Kuramoto model by @chaoming0625 in #15

New Contributors

• @xinzhu-L made their first contribution in #14

Full Changelog: v0.0.3...v0.0.4