MLPro-Int-MuJoCo - Integration of MuJoCo into MLPro

Welcome to MLPro-Int-MuJoCo, an extension to MLPro to integrate the MuJoCo package. MLPro is a middleware framework for standardized machine learning in Python. It is developed by the South Westphalia University of Applied Sciences, Germany, and provides standards, templates, and processes for hybrid machine learning applications. MuJoCo (Multi-Joint dynamics with Contact), in turn, is a physics engine primarily used for simulating and optimizing complex robotic systems and physical interactions in various fields such as robotics, biomechanics, and machine learning.

MLPro-Int-MuJoCo offers wrapper classes that allow the reuse of functionalities from MuJoCo in MLPro.

Preparation

Before running the examples, please install the latest versions of MLPro and MuJoCo as follows:

pip install mlpro-int-mujoco[full] --upgrade

See also

• MLPro - Machine Learning Professional

• MLPro-BF - State-based Systems

• MuJoCo - Advanced physics simulation

• Further MLPro extensions

• MLPro-Int-MuJoCo on GitHub

Example Pool

• Basic Functionalities
  • Howto BF-SYSTEMS-001: Double Pendulum Systems wrapped with MuJoCo
  • Howto BF-SYSTEMS-002: Cartpole Continuous Systems wrapped with MuJoCo
• Reinforcement Learning
  • Howto RL-ENV-001: Run Agent with Random Policy in Native Double Pendulum Environment

API Reference

• Wrapper Root Class
  • CallbacksViewer
  • BaseViewer
  • OffRenderViewer
  • RenderViewer
  • MujocoHandler

About

• Release Notes
• Disclaimer
  • Disclaimers for MLPro Usage
  • Consent