Clay · Noj · Calva — in the browser

Clojure data science,
no setup required.

Hosted Clay notebooks for the scicloj community. Sign in, spin up a notebook, and try real Clojure data science in the browser — Noj loaded and ready to go.

Create your notebook

One free notebook per person while this is a prototype.

A fresh notebook open in the browser: the welcome cell shows a Clojure ns form requiring tablecloth, tableplot, plotje, and fastmath, under the heading "Welcome to your Clojure data science notebook."

What's in the box

A real environment, ready to go.

Everything you need to get started, already wired together and running in the browser.

A real editor

VS Code with Calva and power tools already installed, wired to a live nREPL with Clojure's language server already running.

Re-render on demand

The environment is set up to re-renders the whole notebook with Clay on every save. Or render form-by-form with Calva's Clay integration.

Noj preloaded

Tablecloth, Tableplot, Plotje, Kindly, and the rest of the scicloj stack, already loaded.

Your own sandbox

Each notebook runs in its own Sprite, a stateful Linux machine.

Share read-only

Anyone can view the rendered notebook without needing to log in or install anything.

Zero setup

No local install, no JVM management, no tooling to set up. Just sign in and you're writing Clojure right away.

More information

FAQs

What is this?

This is an experiment. Clojure has a lot of benefits when it comes to working with data, but it can be a bit daunting knowing where to start. Clojure's data science community, Scicloj, has put tremendous effort into making Clojure's data science toolkit easier to use and more approachable in recent years, and this is one more contribution toward that effort -- a way to test how much of the barrier to adoption is really just the setup. If we can remove all of that friction by offering a truly zero-setup environment to play with, does getting started feel easier?

Here you'll get a fully functional, free notebook with Clojure's full data science stack loaded and running on a real JVM, backed by your very own Sprite.

How is it free?

For now, I am subsidizing this first phase of the experiment. It is free thanks to my generous GitHub Sponsors. If it goes well, I will find a way to make it sustainable, but for now I am happy to fund this initial prototype phase to see if this is even useful. If you would like to see this continue, please try it out and let me know what you think. You can email me, or find me in the Clojurians Zulip or Slack. You can also contribute to the project financially by sponsoring me on GitHub.

What are the limitations of this?

This initial demo is limited to 40 users, each getting approximately 50 hours of notebook usage, with up to 10 users active at any one time. These limitations are in place to keep costs in check until I decide what to do with this project, if anything.

Who are you?

I'm Kira Howe (formerly McLean). I gave a talk about Noj and Clojure's data science stack when it was mostly just a dream back at the conj in 2023. So much has happened since then, and I feel like one of the new biggest hurdles to overcome is making Clojure's ecosystem approachable and easy to use. This is one experiment in that direction.

Get started

Your first notebook is waiting.