locosim_jupyter

This repository is a customized setup based on locosim and lab-docker projects. It utilizes configurations from docker-stuffs and linux-stuffs repositories.

Install and Run Locosim Jupyter

1. Install Docker:

   • Linux:

     curl -fsSL https://raw.githubusercontent.com/lomcin/linux-stuffs/master/install/docker -o install_docker.sh
     chmod +x install_docker.sh
     ./install_docker.sh

     • To use NVIDIA GPUs, follow the instructions provided in Installing the NVIDIA Container Toolkit (Installing with Apt , Configuring Docker).

   • Windows:

     1. Install Windows Subsystem for Linux (WSL) by following the Install Linux on Windows with WSL instructions.
     2. Install Docker by following the Turn on Docker Desktop WSL 2 instructions.
     3. Launch Docker Desktop.
     4. Launch WSL.
     5. To navigate to the home directory in WSL, simply run: cd ~
     6. Open port:

        sudo ufw allow 8888
        sudo ufw enable

2. Clone this repository and run prepare.sh:

   git clone https://github.com/gbrlb/locosim_jupyter.git
   cd locosim_jupyter
   ./prepare.sh

Usage

1. In a terminal, navigate to locosim_jupyter, if it is in the home directory:

   cd ~/locosim_jupyter
   

2. Run Docker with:

   • Linux:

     ./run_linux.sh

   • Linux+gpu: You need to have NVIDIA drivers and the NVIDIA Container Toolkit installed already.

     ./run_linux_gpu.sh

   • Windows WSL:

     ./run_wsl.sh

3. Inside the Docker container, execute:

   lab

4. Access the Jupyter Lab interface by opening http://localhost:8888/ in a web browser.

5. Password Token for Jupyter Lab: locosim

6. In the Jupyter Lab, navigate to L1_kinematics.ipynb located at robot_control/lab_exercises/ to access the example exercise.

Troubleshooting

• To ensure compatibility you may need to disable Secure Boot in your BIOS.

• If you encounter an error while running ./prepare.sh related to downloading the robot_control or ros_impedance_controller packages, you can manually clone the repositories as follows:

  1. Clone the robot_control repository:

  git clone https://github.com/gbrlb/robot_control.git -b jupyter "$(pwd)/$USER/ros_ws/src/locosim/robot_control"

  2. Clone the ros_impedance_controller repository:

  git clone https://github.com/gbrlb/ros_impedance_controller.git "$(pwd)/$USER/ros_ws/src/locosim/ros_impedance_controller"

  Then run the Docker script, and inside the Docker container, rebuild ROS as follows:

  source /opt/ros/noetic/setup.bash
  cd $HOME/ros_ws
  catkin build
  source $HOME/ros_ws/devel/setup.bash