Elixir and Jupyter

When using Python I rely heavily on Jupyter (nee iPython) and now that they are pushing multi-language support I am excited to see what other language communities do with it. In this post I will look at setting it up with an Elixir kernel.


First install Jupyer using conda or pip. I recommend using the Anaconda distribution and the conda package manager if possible. If using anaconda, a simple

$ conda install jupyter

should be enough to get you started.

Note: The jupyter binary and its associated commands (jupyter-console, jupyter-notebook etc.) must be on your path otherwise you will get a message such as jupyter: 'notebook' is not a Jupyter command

Next, we need to install and configure the IElixir kernel.

$ git clone https://github.com/pprzetacznik/IElixir.git
$ cd IElixir
$ mix deps.get
$ MIX_ENV=prod mix compile
$ sh install_script.sh

Note: IElixir states it needs 1.1.0-dev but I am running it with 1.0.5 and it seems to work fine. (Just edit the mix.exs file if you need to downgrade.)

Create a notebook

Jupyter provides a number of different interfaces but by far my favorite is the notebook, a HTML interface similar to the Mathematica interface. To create a notebook with the IElixir kernel, run

$ jupyter notebook

Your browser should open the interface (if it doesn't point it at http://localhost:8888). When you select the new menu item, ielxir should be an option

Jupyter Interface

You can now enter any code that you would enter into an iex session, and save that code as a document for reference or editing later.

Jupyter IElixir Kernel


It is very easy to set up Jupyter to use Elixir thanks to the great work by Piotr Przetacznik. Check it out and see what it can do, its definitely a tool worth having in your toolbox.

comments powered by Disqus