PME-toolkit

PME-toolkit v1.2.0

First stable public release of PME-toolkit.

Features

• Full support for:

  • PME
  • PI-PME
  • PD-PME

• Dual implementation:

  • Python (fully functional, CLI-enabled)
  • MATLAB (aligned implementation for validation and benchmarking)

• JSON-driven workflow shared across implementations

• Analytical backmapping to original parametric variables

• Cross-language validation through regression tests

• Reproducible benchmark structure (glider, airfoil)

Installation

pip install pme-toolkit

Documentation

https://cnr-inm-mao.github.io/pme-toolkit/

Notes

This release establishes PME-toolkit as a reproducible and extensible framework for design-space dimensionality reduction in shape optimization.

The Python implementation is now fully operational, while MATLAB remains available for cross-validation and benchmarking purposes.