Using R, python and julia with JupyterLab

The plain command-line interpreters for R, python and julia are useful, but for data analysis work you may want something more interactive. One option for this is to use JupyterLab. JupyterLab works as a server and a web app that talks to this server - so you work through your web browser. This is great when you start plotting and makes it easy to save

Installing JupyterLab

To install it you need the jupyterlab package, and jupyter kernels for each language you want to use. For the purposes of this page I'm going to assume you have activated a conda environment, and we'll use conda (or rather mamba) to install the package. Follow the tabs below to install what's needed:

1. Install the jupyterlab package using conda:

mamba install jupyterlab

2. Then install the jupyter kernels for your favourite languages.

Installing the Jupyter language kernels

Install the python kernel

To install a python kernel, install the ipython package:

mamba install ipython

(However this is probably already installed.)

Install the R kernel

To do this requires installing the IRkernel package. From a running R session type:

install.packages( 'IRkernel' )
IRkernel::installspec()

Install the Julia kernel

This involves installing the `IJulia` package. From a running julia session, type `]` to get into package mode and type:

add IJulia

Trying JupyterLab

To get started, go to your terminal and run

jupyter lab

This will start the JupyterLab server (which runs in your terminal) and after a few moments will launch a new browser window/tab to interact with it. It will look something like this:

Note

If it doesn't start a browser, do this manually by copying the URL shown in the jupyter lab output and opening it in your browser.

Now you can start a new python, R, or julia 'notebook' session by clicking on the relevant icon.

Exercise Try an R notebook session for example, and let's plot the density of a beta distribution. Type these lines one at a time and press <Shift-Enter> to evaluate each cell:

x = seq( from = 0, to = 1, by = 0.001 )

y = dbeta( x, shape1 = 2, shape2 = 4 )

plot( x, y, type = 'l' )

It should look something like this:

In general notebooks are great for interactive analysis and for generating data-driven reports.

(Jupyter Lab isn't the only way to get this type of functionality - for example, similar functionality can be had using R Markdown documents in RStudio, and for Julia the Julia extension for Visual Studio Code works great.)