Philippa Markovics
David Schmudde2019 Round Up 🍾
It has been an exciting year at Nextjournal. Our users have done a lot of great work since our launch in May and we had some exciting opportunities along the way. Here are the highlights:
Reproducible Research
🦠 We partnered with Will Ludington from Carnegie Science Institute who published all their analysis behind his PNAS Paper Microbiome interactions shape host fitness as runnable notebooks.
🚀 At Curry On 2019, Heather Miller introduced us to Jan Vitek who is one of the main authors of the paper On the Impact of Programming Languages on Code Quality. The paper shows a correlation between specific languages like Java, Clojure, Haskell, and C and the quality of code programmers produced. There was one problem — the results could not be reproduced.
This encounter led to hosting the paper’s replication attempt as reproducible artifact on Nextjournal. For an introduction to the general issue read Walk the Walk: Reproducibility All the Way Down.
🛸 James Winters published Cumulative Culture in Open-Ended Problem Spaces which offers an interesting approach to modeling. Remix it to do some experimentation yourself!
ECOOP 2020
🏛 Meeting Jan Vitek also led to an opportunity we are incredibly excited about: In 2020 we will partner with the European Conference on Object-Oriented Programming (ECOOP) and allow applicants to submit their artifacts as runnable notebooks!
Executable Books: Programming Language Foundations in Agda
📜 Philip Wadler gave us permission to port his book Programming Language Foundations in Agda to Nextjournal. (Another great thing that came out of Curry On 2019.) Remix the individual notebooks and try the exercises yourself!
Runnable Tutorials
⏳ Juxt created a story-based introduction course to Crux that you can experiment with in A Bitemporal Tale – History of Histories. They also provide their entire series of Crux tutorials as Nextjournal notebooks.
🌀 Chris Rackauckas uploaded his workshop Solving Partial Differential Equations (PDEs) in Julia. Remix, run, and learn Julia!
📈 All R and Python guides for Facebook’s Prophet library have been published on Nextjournal as well as Dr. Eric Brown’s four part series on Forecasting Time Series Data With Prophet. Remix and run the docs to experiment with one of the most popular forecasting libraries around!
Teaching
We’re also happy to see that educators are using Nextjournal in the classroom:
Frederic Peschanski uses Clojure and his LaTTe proof assistant library to formalize the proof of the Knaster-Tarski fixed point theorem.
Nick Doiron gave a workshop on Geospatial API with Python and PostGIS at PyCon India.
If teaching with interactive notebooks is something you’re interested in, please don’t hesitate to get in touch!
Julia Seasons of Contributions
The Julia Seasons of Contributions (JSoC) are the seasonal programs for funding students and other developers to contribute to the open source ecosystem. Many of the participants used Nextjournal as a blog to publish updates and runnable code during their tenure. You can check out all 40 (!) notebooks in the JSOC notebook collection. Highlights include:
🦾 Ludovico Bessi explores different ways of using machine learning and surrogate models to solve problems with costly computations. Get to know his new Julia library,
surrogates.jl, through the runnable notebooks he kept during its development🕸 Akshay Jain details their JSoC contribution to NumFOCUS open tools for science. The
TraceEstimation.jllibrary will help Julia coders work more efficiently in domains such as biology, chemistry, physics, machine learning, and network theory.
More Highlights
🌵 Cactus to Clouds processes 20 years of seismic data in the cloud using Julia and Amazon Web Services.
🐟 Multilevel Regression for Sockeye Salmon Spawner Abundance Across Years and Rivers — Collecting accurate data in complex natural habitats is difficult. Scott Akenhead takes a look at different models that recover parameters related to sockeye salmon spawning across various rivers and time spans.
💸 Chad Scherrer explores optimal pricing using Bayesian statistical methods.
🧮 Association Rule Mining offers another method to identify novel patterns in large sets of data. This collection of notebooks provides an overview of what it's all about and how it can be applied.
The Nextjournal Team
We spent most of our time after launch improving the notebook experience for our users. But we were also using Nextjournal ourselves to talk about compiler design, art, declarative programming, explorable explanations, and the reproducibility crisis itself.
Conferences
In May, Martin gave a talk about Nextjournal’s transition from Elm and Elixir to Clojure at ClojureD in Berlin. Watch the video here.
We also had the opportunity to be sponsors at Curry On 2019. Having a booth there and be visible was a great way to connect with the Computer Science community and better understand their needs. Thanks to Heather Miller for making it happen!
Simon and David both presented at Curry On using Nextjournal’s experimental slideshow mode:
David’s talk Say What You Mean - Exploring Declarative Computation in Art included code cells that demonstrated experimental notebook self-modification. Live code! Here’s the video and the remixable notebook.
In his talk Julia: A Compiler for the Future, Simon shows how the Julia language combines the best of Python and C++, in both compiler technology and language design. Watch the video here.
They also both presented at PyData Berlin:
Simon gave an introduction to Julia. Check out Julia for Pythonistas.
David gave an intro to functional programming techniques in Python. Check out Dr. Schmood’s Notebook of Python Calisthenics and Orthodontia.
Simon had one more at JuliaCon in Lisbon where he talked about PackageCompiler for the Julia language. Watch the video here.
In July, we also attended ECOOP 2019 in Athens to further connect with Computer Science researchers and understand their needs.
Scicloj
Martin took a Nextjournal deep dive in Improving the Data Science Workflow with Immutability at Scicloj, a community of people creating high-quality open source solutions for data science in Clojure.
The Green Web Foundation
Chris Adams spent a day with us at our office in Berlin to better understand our carbon emissions and figuring out how we can become carbon-neutral. This was another great thing that came out of visiting the Heart of Clojure conference. Be sure to watch Chris’ talk: What you can do when software is heating the world?
Blog Posts
⚗️ Philipp wondered where the tools are to create explorable explanations in Towards Model-Driven Authoring: What’s Missing?
🎨 The Shape of Intuition mixes the work of Jackson Pollock, Anna Teresa De Keersmaeker, and Steve Reich with explorable code to assert computation’s presence in instinctual processes.
⏰ Martin addressed our core mission in No, your ten rules aren’t simple and they won’t fix Reproducibility and followed it up with solutions in Towards Reproducibility: Git.
Podcast
Finally, Martin was interviewed in December by Ray McDermott and Vijay Kiran at the defn podcast. Their conversation includes Schafkopf, Nextjournal, transitioning from Elm and Elixir to Clojure and ClojureScript and much more.
We are incredibly grateful for all the fantastic content that was produced in 2019 and are looking forward to some exciting things coming in 2020! Thanks everyone and have a happy new year! 🙏