Machine: Ubuntu 20.04.3 LTS focal
conda create --name offlineRL python=3.8
conda activate offlineRL
conda install pip
git clone https://github.com/rail-berkeley/d4rl.git
cd d4rl
pip install -e .
cd ..
wget https://mujoco.org/download/mujoco210-linux-x86_64.tar.gz
tar -xvzf mujoco210-linux-x86_64.tar.gz
mkdir ~/.mujoco
mv mujoco210 ~/.mujoco/
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/ivan/.mujoco/mujoco210/bin' >> ~/.bashrc
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib/nvidia' >> ~/.bashrc
. ~/.bashrc
mkdir CARLA
cd CARLA
wget https://carla-releases.s3.eu-west-3.amazonaws.com/Linux/CARLA_0.9.8.tar.gz
wget https://carla-releases.s3.eu-west-3.amazonaws.com/Linux/AdditionalMaps_0.9.8.tar.gz
tar -xvzf CARLA_0.9.8.tar.gz
tar -xvzf AdditionalMaps_0.9.8.tar.gz
echo 'export PYTHONPATH=$PYTHONPATH:/home/CARLA_0.9.8/PythonAPI' >> ~/.bashrc
echo 'export PYTHONPATH=$PYTHONPATH:/home/CARLA_0.9.8/PythonAPI/carla' >> ~/.bashrc
echo 'export PYTHONPATH=$PYTHONPATH:/home/CARLA_0.9.8/PythonAPI/carla/dist/carla-0.9.8-py3.5-linux-x86_64.egg' >> ~/.bashrc
conda activate offlineRL
conda install -c anaconda networkx
pip install pygame
conda install -c conda-forge dotmap
cd ..
git clone https://github.com/flow-project/flow.git
cd flow
Error without it: redis 2.10.6 is installed but redis>=3.3.2 is required by {‘ray’}
I repleace the “redis~=2.10.6” by “redis” in enviroment.yml and requirements.txt.
conda env create -f environment.yml
conda activate flow
python setup.py develop
Next, we install the SUMO.
sudo apt-get install sumo sumo-tools sumo-doc
Checking
which sumo
sumo --version
sumo-gui
PS: It still does’t work
I got errors, so I needed to install those.
sudo apt install libx11-dev
sudo apt install libglew-dev
sudo apt-get install patchelf
sudo apt install libosmesa6-dev
conda activate offlineRL
Write to file test_env.py:
import gym
import d4rl # Import required to register environments
# Create the environment
env = gym.make('maze2d-umaze-v1')
# d4rl abides by the OpenAI gym interface
env.reset()
env.step(env.action_space.sample())
# Each task is associated with a dataset
# dataset contains observations, actions, rewards, terminals, and infos
dataset = env.get_dataset()
print(dataset['observations']) # An N x dim_observation Numpy array of observations
# Alternatively, use d4rl.qlearning_dataset which
# also adds next_observations.
dataset = d4rl.qlearning_dataset(env)
run:
python test.py
All matireals from: