Installation

Jupyter-ROS is distributed as a conda, pip, and npm package. We recommend installing it inside a conda environment. To install conda, we suggest using the Miniconda installer.

# Option 1. conda or mamba [Recommended]
$ conda create -n jupyros_env python=3.9
$ conda activate jupyros_env
$ conda install jupyter-ros -c robostack

# Option 2. pip
$ pip install jupyros

# Option 3. npm
$ npm i jupyter-ros

The jupyter-ros server extension

The jupyter-ros package contains an optional server extension which can be used to serve static files from a catkin workspace to the web browser. For the 3D widgets we might have to load meshes to display robots correctly. These meshes are usually part of a robot description package in ROS. In order to enable the web browser to access those meshes, we can enable the jupyter-ros server extension:

$ jupyter serverextension enable jupyros

Enabling: jupyros
- Writing config: /home/user/.jupyter
    - Validating...
      jupyros 0.6.0 OK

The jupyter-ros server extension adds a handler to the Jupyter server that will return contents from ROS packages. To check that it works, you can (re-)start the Jupyter server and navigate to localhost:8888/rospkg/rospy/package.xml or any other file that should exist in your catkin workspace.

Warning

There are currently no security features in the jupyter-ros server extension. If you have sensitive ROS packages, all contents (including uncompiled source code) can be found through this extension.