Julia Comes to Google Colab – TheCoatlessProfessor

As a long-time Julia lurker, I was thrilled to discover that Google Colab now offers native Julia support through IJulia! The work to incorporate Julia into Colab was led by Eric Johnson (metrizable) and means that Colab officially is Jupyter-complete, supporting the triad of languages that make up Jupyter’s name: Julia, Python, and R.

You can try it out now through a practical Julia Colab notebook that I’ve created.

Or, you can create a new Julia notebook in Colab by:

Clicking this link: https://colab.research.google.com/notebook#create=true&language=julia
Inside of Colab, click File -> New notebook in Drive and, then, in the new notebook select Runtime -> Change runtime type -> Julia.

If you’re curious about the new Julia support and why it matters, read on for more details.

Farewell to Hackway Workaround Scripts

For those who have been using Julia with Colab, you’ll remember the old workflow that required using a template notebook with a shell script installation:

%%shell
set -e

JULIA_VERSION="1.8.2" # any version ≥ 0.7.0
JULIA_PACKAGES="IJulia BenchmarkTools"
# ... more configuration and installation code ...

While this solution worked, it required page reloads, occasional runtime resets, and repeating the setup process after inactivity timeouts. Not ideal when you’re in the middle of exploring data or developing algorithms!

With native support, we can simply select Julia as our runtime environment and start coding immediately. No more installation scripts, no more waiting for packages to compile on every session, and no more workarounds.

This development is particularly exciting for:

Teachers and students who can now easily incorporate Julia into computational courses
Researchers who want to leverage Julia’s speed for data analysis without local installation
Curious programmers looking to experiment with Julia’s multiple dispatch paradigm
Data scientists interested in Julia’s growing ecosystem of statistical and ML packages

Exploring Julia in Colab

To celebrate, I’ve created a practical Julia Colab notebook that walks through key Julia features specifically in the Colab environment. The notebook covers:

Colab-specific Julia features and capabilities
Julia language overview with comparisons for R and Python users
Data wrangling and visualization

If you’ve been curious about Julia but hesitated due to setup friction, this native Colab support removes that barrier. You can now experiment with Julia directly in your browser without any local installation.

You can also find the notebook on GitHub if you want to contribute improvements or suggest additions.

Fin

This is an exciting step forward for Julia’s accessibility in cloud environments. I’m particularly interested to see how the community leverages the GPU and TPU access for machine learning and scientific computing applications.