setup_opencda.txt

# setup steps

# activate conda in shell
eval "$(/opt/miniconda3/bin/conda shell.bash hook)"

# go to OpenCDA folder
cd /opt/OpenCDA/

# create your own conda environment
conda env create -f environment.yml

conda activate opencda

# install dependencies
conda install pytorch==1.8.0 torchvision==0.9.0 torchaudio==0.8.0 cudatoolkit=11.8 -c pytorch -c conda-forge

pip install poetry

poetry install

run the following
----
CARLA_HOME=/opt/carla/
CARLA_VERSION=0.9.12
CARLA_EGG_FILE=${CARLA_HOME}/PythonAPI/carla/dist/carla-"${CARLA_VERSION}"-py3.7-linux-x86_64.egg
CACHE=${PWD}/cache

mkdir -p "$CACHE"
cp $CARLA_EGG_FILE $CACHE

unzip "${CACHE}"/carla-"${CARLA_VERSION}"-py3.7-linux-x86_64.egg -d "${CACHE}"/carla-"${CARLA_VERSION}"-py3.7-linux-x86_64

SETUP_PY=${PWD}/scripts/setup.py
cp "$SETUP_PY" "${CACHE}"/carla-"${CARLA_VERSION}"-py3.7-linux-x86_64/

# should already be in the conda environment from before
pip install -e ${CACHE}/carla-"${CARLA_VERSION}"-py3.7-linux-x86_64
----

then re-run this

conda install pytorch==1.8.0 torchvision==0.9.0 torchaudio==0.8.0 cudatoolkit=11.8 -c pytorch -c conda-forge


# run simulation

first run carla sim in /opt/carla which is CarlaUE4.sh

then run an example from OpenCDA i.e. 

python opencda.py -t platoon_joining_2lanefree_carla -v 0.9.12 --apply_ml

listof examples in opencda/scenario_testing/