TORCS Simulator and Custom Environments

Overview

This repository includes the TORCS race simulator (version 1.3.7) with minimal modifications, as well as customized gym environments compatible with Gymnasium.

TORCS Simulator: Located in src/torcs, this is version 1.3.7 of the TORCS race simulator with a few minimal changes.
Gym Torcs Environment: Located in src/gymnasium_torcs, this version includes changes to make it compatible with Gymnasium.
Custom TORCS Environment: Located in src/torcs_lidar_environment, this environment uses 19 lidar rays for observations and the steer angle as the action.

Installation

To set up the project, follow these steps:

1. Clone the repository:

git clone https://github.com/yourusername/torcs_simulator.git
cd torcs_simulator

2. Create conda environment:

conda create -n torcs_gym python==3.11

3. Install poetry:

pip install poetry

4. Install dependencies:

python poetry_install.py

5.Build and install TORCS:

bash install_torcs.sh

Running with Docker

To run this project using Docker, make sure you have Docker installed on your system. If not, you can download and install Docker from Docker's official website.

Build the Docker image:

Note: If you already build torcs outside docker please clean the torcs project as it is a bit dirty, and it will interfere with torcs building in docker:
```
cd src/torcs
rm -rf BUILD
make clean
```
Navigate to the root directory of the project where Dockerfile is located, then run:
```
docker build -t torcs-simulator .
```
and run docker using:
```
docker run -it --rm --privileged --net=host \
   --env DISPLAY --volume /tmp/.X11-unix:/tmp/.X11-unix \
   torcs-simulator bash
```
In case you experience some issues with the rendering when using docker make sure to add the docker user to xhost. So run on your local machine:
```
xhost +SI:localuser:docker_user
```

Usage

from gymnasium_torcs.gym_torcs import TorcsEnv

env = TorcsEnv()
observation = env.reset()
done = False

while not done:
    action = env.action_space.sample()  # Replace with your action
    observation, reward, done, truncations, info = env.step(action)

env.end()

Contributing

Feel free to submit issues or pull requests if you have suggestions or improvements.

License

This project is licensed under the MIT License. See the LICENSE file for details.

ivandrodri/torcs_gymnasium