This is the public code repository for the paper "Maximum likelihood constraint inference on continuous state spaces," presented at the ICRA 2022 conference.
Ubuntu 20.04 and Python 3.8 is recommended. First, make sure required system packages are installed:
sudo apt install mpich swig
Next, install required Python packages (consider making a virtual environment first with python -m venv venv_MLCI):
pip install -r requirements.txt
pip install -e custom_envs
python ppo_pendulum.py --logdir logs/pendulum --experiment 1 --entropy 1.5 --numEpochs 40
python do_rollouts.py logs/pendulum --runExperiments startsEnds5.pkl --singleExperiment 1 --samplesPerExperiment 10000 --outputFile data/pendulum/expected_rollouts.pkl
python contraints_from_trajectories.py --sampleFile data/pendulum/expected_rollouts.pkl --outputFile data/pendulum/expected_constraints.pkl --envType pendulum
python constraints_from_trajectories.py --demoFileBase logs/pendulum/pendulumAgentShort_NoLowVeloAt5Tenths --outputFile data/pendulum/demo_constraints.pkl --envType pendulum