Reflections on useR! 2014

UseR! 2014, the R user conference held last week in LA, was the most successful yet. Around 700 R users from around the world converged on the UCLA campus to share their experiences with the R language and to socialize with other data scientists, statisticians and others using R.

The week began with a series of 3-hour tutorials on topics as diverse as data management, visualization, statistics and biostatistics, programming, and interactive applications. (Joe Rickert reported on the tutorials from the field last week.) Unlike in previous years, the tutorial day was included in the registration, which meant that all of the sessions were jam-packed with interested R users.

The remaining 2-and-a-half days were packed with keynotes, contributed talks, and social sessions galore. Given the parallel nature of the tracks I couldn't make it to more than a fraction of the talks, but here are a few of the highlights from my notes:

• In the opening keynote, John Chambers shared the story of the genesis of the S language (which later begat R — see more in this 2013 interview). The three key principles behind the S language were objects (everything is an object), functions (everything that happens is a function call), and interfaces (the language is an interface to other algorithms). In fact, the very first sketch of the language (from a slide made in 1976, shown below), called described it as an "Algorithm Interface".

• Jeroen Ooms demonstrated OpenCPU, a Web API for scientific computing. OpenCPU makes it possible to integrate R into web-based apps for non R users. You can see some examples of OpenCPU in action in the App Gallery, including the Stockplot app that Jeroen demonstrated.

• Karthik Ram talked about the ROpenSci project and fostering open science with R. The ROpenSci team has created dozens of R packages, including interfaces to public data sources, data visualization tools, and support for reproducible research. 

• Google has more than 1000 R users on its internal R mailing list, according to Tim Hesterberg in Google's sponsor talk. (My sponsor talk for Revolution Analytics can be found here.)

• The heR panel discussion and mixer, which facilitated an excellent conversation about women in data science and the R community. (The useR! conference itself was around 25% women — certainly room for improvement, but better than many math or computer science conferences.)

• Thomas Fuchs from NASA/JPL, who revealed in two talks that R is used for vision analysis in space exploration (including the Mars Hirise mission and deep-space astronomy). As a NASA buff this was a thrill for me to learn, and I hope to write more about it sometime.

• The conference banquet under the summer twilight on the UCLA lawns and featuring entertaining acecdotes from David McArthur.

For more on the conference, the slides from many of the talks and tutorials are available at the useR! website. (If you presented there, please submit your slides via a pull request.) Also check out these reviews of the conference from Daniel Gutierrez at InsideBigData and Phyllis Zimbler Miller.

The next conference, useR! 2015 will be held in Aarlborg, Denmark. I'm looking forward to it already!

Comments

[Ari]

David,

I attended useR! for the first time this year and had a great time. I work as a software engineer and am a big proponent of R within my company. We have lots of data analysis needs and R is great for that.

I found the tutorials and talks by the RStudio folks were particularly relevant to me - they seem to be doing a lot of work to make R more intuitive and accessible to people who develop in more traditional languages.

Also, it was nice to finally meet you after reading your blog for so long!

Ari