Implementation:Farama Foundation Gymnasium InvertedPendulumEnv V5

Knowledge Sources	Farama_Foundation_Gymnasium Gymnasium Docs
Domains	Reinforcement_Learning, MuJoCo_Environments
Last Updated	2026-02-15 03:00 GMT

Overview

Concrete implementation of the InvertedPendulum v5 MuJoCo environment provided by Gymnasium.

Description

The InvertedPendulum v5 environment is the Cartpole environment powered by MuJoCo, based on Barto, Sutton, and Anderson's work in "Neuronlike adaptive elements that can solve difficult learning control problems". It consists of a cart that can be moved linearly, with a pole attached to one end. The goal is to balance the pole on top of the cart by applying forces. The observation space is 4-dimensional (cart position, pole angle, cart velocity, pole angular velocity). The reward is +1 for each timestep the pole remains upright (angle less than 0.2 radians). The episode terminates if any observation is non-finite or the pole angle exceeds 0.2 radians.

Usage

Use this environment as a simple MuJoCo control benchmark. It serves as a good starting point for verifying RL algorithm implementations before moving to more complex environments. The v5 version adds configurable reset_noise_scale, custom xml_file support, and reward info.

Code Reference

Source Location

Repository: Farama_Foundation_Gymnasium
File: gymnasium/envs/mujoco/inverted_pendulum_v5.py

Signature

class InvertedPendulumEnv(MujocoEnv, utils.EzPickle):
    def __init__(
        self,
        xml_file: str = "inverted_pendulum.xml",
        frame_skip: int = 2,
        default_camera_config: dict[str, float | int] = DEFAULT_CAMERA_CONFIG,
        reset_noise_scale: float = 0.01,
        **kwargs,
    )

Import

import gymnasium as gym
env = gym.make("InvertedPendulum-v5")

I/O Contract

Inputs

Name	Type	Required	Description
action	np.ndarray (1,)	Yes	Force applied on the cart, range [-3, 3]

Outputs

Name	Type	Description
observation	np.ndarray (4,)	State vector: cart position, pole angle, cart velocity, pole angular velocity
reward	float	1.0 if not terminated, 0.0 otherwise
terminated	bool	True if observation is non-finite or pole angle exceeds 0.2 radians
truncated	bool	Episode truncation (handled by TimeLimit wrapper, default 1000 timesteps)
info	dict	Contains reward_survive

Usage Examples

import gymnasium as gym

env = gym.make("InvertedPendulum-v5")
observation, info = env.reset(seed=42)

for _ in range(1000):
    action = env.action_space.sample()
    observation, reward, terminated, truncated, info = env.step(action)
    if terminated or truncated:
        observation, info = env.reset()

env.close()

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment