ParMOO: Python library for parallel multiobjective simulation optimization
ParMOO is a Python library for solving multiobjective simulation-based optimization problems, while exploiting problem structure.
ParMOO stands for “parallel multiobjective optimization”. ParMOO can be used to solve multiobjective optimization problems (MOOPs) or to generate batches of simulation inputs for parallel evaluation.
ParMOO is intended for scientists, engineers, optimizers, and other practitioners, who are looking to build or use custom solvers for computationally expensive multiobjective problems.
As of version 0.4.0: we are integrating with jax!
If you’re new to ParMOO:
Check out the Quickstart
Try some of our Basic Tutorials
Check us out on GitHub
Table of Contents
User Guide:
API:
- ParMOO API
- Basic ParMOO Classes/Objects
- Embedder Classes for Custom Variable Types
- SimulationDatabase Classes for ParMOO’s Multiobjective Database
- Acquisition Functions
- Surrogate Optimizers
- Search Techniques
- Surrogate Functions
- Built-in Problem Libraries and Custom Problem Templates
- The DTLZ Problem Library
- Extra Developer Tools
- The Interactive Visualization (viz) Library
Tutorials:
Developer Guide: