L.A. Unconf-idential : a.k.a. an rOpenSci #runconf17 Retrospective

Last year, I was able to sit back and lazily “RT” Julia Silge’s excellent retrospective on her 2016 @rOpenSci “unconference” experience. Since Julia was not there this year, and the unconference experience is still in primary storage (LMD v2.0 was a success!) I thought this would be the perfect time for a mindful look-back.

And Now, A Word From…

Hosting a conference is an expensive endeavour. These organizations made the event possible:

At most “conferences” you are inundated with advertising from event sponsors. These folks provided resources and said “do good work”. That makes them all pretty amazing but is also an indicator of the awesomeness of this particular unconference.

All For “Un” and “Un” For All

Over the years, I’ve become much less appreciative of “talking heads” events. Don’t get me wrong. There’s great benefit in being part of a larger group experiencing the same message(s) and getting inspired to understand and investigate new ideas, concepts and technologies. Shining examples of what great “conferences” look like include OpenVis Conf and RStudio’s inaugural self-titled event.

The @rOpenSci “unconference” model is incredibly refreshing.

It has the “get’er done” feel of a hackathon but places less importance on the competitive aspect that is usually paramount in hackathons and increases emphasis on forging links, partnerships and creativity across the diverse R community. It’s really like the Three Musketeers saying “all for one and one for all” since we were all there to help each other build great things to enable R users to build even greater things.

What We Going To Do Tonight, BrainKarthik?

I’ll let you peruse the rOpenSci member list and #runconf17 attendee list at your leisure. Those folks came to Los Angeles to work — not just listen — for two days.

In the grand scheme of things, two days is not much time. It takes many organizations two days to just agree on what conference room they’re going to use for an upcoming internal meeting let alone try to get something meaningful accomplished. In two days, the unconference participants cranked out ~20 working projects. No project had every “i” dotted and every “t” crossed but the vast majority were at Minimum Viable Product status by presentation time on Day 2, and none were “trivial”.

You can read all of the projects at the aforementioned link. Any that I fail to mention here is not a reflection on the project but more a factor of needing to keep this post to a reasonable length. To that end, I’m not even elaborating on the project I mainly worked on with Rich, Steph, Oliver & Jeroen (though it is getting a separate blog post soon).

Want to inspire Minecraft enthusiasts to learn R? There’s an app for that. The vast functional programming power that’s enabled the modern statistics and machine learning revolution is now at the fingertips of any player. On the flip side, you now have tools to create 3D models in a world you can literally walk through — as in, literally stand and watch models of migratory patterns of laden swallows that you’ve developed. Or, make a 3D scatterblock™ diagram and inspect — or destroy with an obsidian axe — interesting clusters. Eliminating data set outliers never felt so cathartic! Or, even create mazes algorithmically and see if your AI-controlled avatar can find its way out.

Want to connect up live sensor (or other live stream) data into an R Shiny project? There’s an app for that. Websockets are a more efficient & versatile way to wire up clients and servers. Amazon’s IoT platform even uses it as one way to push data out from your connected hairbrush. R now has a lightweight way to grab this data.

The team even live-demoed how to pick up accelerometer data from a mobile device and collect + plot it live.

Want vastly improved summaries of your data frames so you can find errors, normalize columns and get to visualization and model development faster? There an app for that.

Yes, I — too — SQUEEd at in-console & in-data frame histograms.

There are many more projects for you to investigate and U.S. folks should be thankful for a long weekend so they have time to dive into each of them.

It’s never about the technology. It’s about the people.

(I trust Doctor Who fans will forgive me for usurping Clara’s best line from the Bells of Saint John)

Stefanie, Karthik, Scott & the rest of the rOpenSci team did a phenomenal job organizing and running the unconference. Their efforts ensured it was an open and safe environment for folks (or ?) to just be themselves.

I got to “see” idividuals I’ve only ever previously digitally interacted or collaborated with. Their IRL smiles — a very familiar expression on the faces of attendees during the two-day event — are even wider and brighter than those that come through in their tweets and blog posts.

Each and every attendee I met brought fresh perspectives, unique knowledge, incredible talent and unwavering enthusiasm to the event. Teams and individuals traded ideas and code snippets and provided inspiration and encouragement when not hammering out massive quantities of R code.

You can actually get a mini-unconf experience at any time from the comfort of your own glowing rectangle nesting spot. Pick or start a project, connect with the team and dive in.

FIN

It was great meeting new folks, hanging with familiar faces and having two days to just focus on making things for the R community. I hope more conferences or groups explore the “un” model and look forward to seeing the 2017 projects become production-ready and more folks jumping on board rOpenSci.