Tested on Ubuntu 22.04 (arm64 and x64)
- Install dependicies
- pip:
sudo apt install python3-pip - colcon:
pip install -U colcon-common-extensions
| from navground import sim | |
| from navground.learning import ControlActionConfig, DefaultObservationConfig, GroupConfig | |
| from navground.learning.parallel_env import parallel_env, make_vec_from_penv | |
| from navground.learning.rewards import SocialReward | |
| from stable_baselines3 import SAC | |
| # 1. Load a scenario | |
| scenario = sim.load_scenario(...) | |
| # 2. Create a training environmnent |
| steps: 600 | |
| time_step: 0.03 | |
| scenario: | |
| type: CrossTorus | |
| side: 10 | |
| groups: | |
| - type: wheelchair | |
| color: darkred | |
| number: 4 | |
| radius: 0.25 |
sudo apt-get install libopencv-dev python-opencvInstall ROS kinetic (ros-kinetic-desktop-full)
http://wiki.ros.org/ROS/Tutorials/InstallingandConfiguringROSEnvironment