Improve this page

Basic interface

All environments share two fundamental functions reset() and step(action). reset restarts the environment to the initial state. step sends muscle excitations and runs the simulation for one step. These two functions are the minimal requirement for most of the reinforcement learning algorithms.

Initialization

You can see all available environments in this document. Here we describe how to initialize and run the most basic one Arm2DEnv

In order to create an environment, use:

from osim.env import Arm2DEnv
env = Arm2DEnv(visualize = True)

Parameters:

  • visualize - turn the visualizer on and off

Methods

reset(project = True)

Restart the environment to the initial state. Note that extra parameters can be available depending on the environment.

The function returns:

  • observation - a vector (if project = True) or a dictionary describing the state of muscles, joints, and bodies in the biomechanical system.

step(action, project = True)

Make one iteration of the simulation.

  • action - a list of numbers in the [0,1] interval, corresponding to excitations of muscles exposed in the environment.

The function returns:

  • observation - a vector (if project = True) or a dictionary describing the state of muscles, joints, and bodies in the biomechanical system.

  • reward - reward gained in the last iteration.

  • done - indicates if the move was the last step of the environment.

  • info - for compatibility with OpenAI gym, currently not used.