This is the code associated with the research project Interpreting Contact Interactions to Overcome Failure in Robotic Assembly Tasks
Run the following lines on a linux terminal to install the required software libraries to run the paper code, please pay attention to the comments above each set of lines to run to make sure you are using the right paths for your file system.
#Assuming you have installed Anaconda 3
conda create -n sens_search python=3.7
conda activate sens_search
#Setting environmental variables for running mujoco and activating renderer
cd $CONDA_PREFIX
mkdir -p ./etc/conda/activate.d
mkdir -p ./etc/conda/deactivate.d
touch ./etc/conda/activate.d/env_vars.sh
touch ./etc/conda/deactivate.d/env_vars.sh
#CHECK PATHS TO MUJOCO make sure these paths are correct
echo '#!/bin/sh' >> ./etc/conda/activate.d/env_vars.sh
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/scr-ssd/software_packages/mujoco/mujoco200/bin' >> ./etc/conda/activate.d/env_vars.sh
echo 'export MUJOCO_PY_MUJOCO_PATH=/scr-ssd/software_packages/mujoco/mujoco200/' >> ./etc/conda/activate.d/env_vars.sh
echo 'export MUJOCO_PY_MJKEY_PATH=/scr-ssd/software_packages/mujoco/mjkey.txt' >> ./etc/conda/activate.d/env_vars.sh
source ./etc/conda/activate.d/env_vars.sh
echo '#!/bin/sh' >> ./etc/conda/deactivate.d/env_vars.sh
echo 'unset LD_LIBRARY_PATH' >> ./etc/conda/deactivate.d/env_vars.sh
echo 'unset MUJOCO_PY_MUJOCO_PATH' >> ./etc/conda/deactivate.d/env_vars.sh
echo 'unset MUJOCO_PY_MJKEY_PATH' >> ./etc/conda/deactivate.d/env_vars.sh
#CHECK TO MAKE SURE ITS THE RIGHT BASH FILE make sure this alias is going to the right bash file either bashrc.user or create a bash_aliases
echo "alias onscreen_render_muj='export LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libGLEW.so'" >> ~/.bash_aliases
echo "alias offscreen_render_muj='unset LD_PRELOAD'" >> ~/.bash_aliases
#Installing pytorch make sure you have the right cudatoolkit for your computer
conda install pytorch torchvision cudatoolkit=9.2 -c pytorch
conda install -c conda-forge tensorboard
conda install -c conda-forge tensorboardx
conda install -c intel scikit-learn
#echo 'export USE_DAAL4PY_SKLEARN=YES' >> ./etc/conda/activate.d/env_vars.sh
#echo 'unset USE_DAAL4PY_SKLEARN' >> ./etc/conda/deactivate.d/env_vars.sh
conda install -c conda-forge gym
conda install h5py
conda install matplotlib
#CHECK PATH create project directory
mkdir /scr-ssd/sens_search/
cd /scr-ssd/sens_search/
#Adding and installing robosuite
git clone https://github.com/ARISE-Initiative/robosuite
pip install hjson pyquaternion pyyaml mujoco-py pybullet==1.9.5 gtimer
sudo apt install libosmesa6-dev libgl1-mesa-glx libglfw3
cd robosuite
pip install -e .
cd ..
#git clone -b https://github.com/zachares/rlkit.git
#cd rlkit
#pip install -e .
#cd ..
#git clone https://github.com/vitchyr/viskit.git
#cd viskit
#pip install -e .
#cd ..
git clone -b paper_code https://github.com/zachares/deep_learning.git
cd deep_learning
pip install -e .
cd ..
git clone -b sens_search https://github.com/zachares/sensorimotor_search.git
cd sensorimotor_search
pip install -e .
cd ..