Skip to content

VAMOS 1.0.1 release notes

Released 2026-10-01.

VAMOS 1.0.1 packages the fixes and documentation updates merged after the first public release. The stable 1.x API and the 1.0.0 run and study schemas remain the compatibility baseline.

Algorithm and evaluation fixes

  • NSGA-II applies the requested steady-state configuration, including with the Numba kernel.
  • SMPSO honors population and non-dominated result modes and stores an explicitly configured external result archive separately from its internal leaders. Result storage does not change leader selection or swarm dynamics.
  • AGE-MOEA and RVEA apply their configured constraint handling.
  • Experimental Dask evaluation can discover the active client. Serial fallback still requires explicit opt-in.

Tuning and documentation

CLI tuning scores a consistent final-population source, validates constrained algorithm support before execution, and namespaces persistent Optuna studies for the corrected scoring contract. This experimental tuning change can alter selected configurations; record the package version with scientific results.

The documentation includes verified installation and troubleshooting recipes, algorithm and problem decision tables, MOEA/D and NSGA-III tutorials, and real result plots. The retired multilingual site and internal audit and architecture decision pages have been removed. Canonical contributor contracts remain in Architecture Health and the linked guides.

Installation and compatibility

python -m pip install --upgrade "vamos-optimization==1.0.1"

Python 3.10 or newer remains required. The 1.0.0 compatibility fixtures and schemas are preserved. Exact replay requires a matching material environment; an upgrade is not a promise of identical trajectories for algorithms whose behavior was corrected. Keep the original environment when replaying an older recorded experiment.

See stability and versioning, known limitations, and release verification.