rstudio::conf + workshop 2020

Image

I represented our group at rstudio::conf 2020 during the last week of January. This conference is hosted by the organization responsible for the RStudio development environment and for many other improvements in the useability of the R programming language over the past decade, including packages and training. The conference was located in San Francisco this year, and there were more than 2,000 attendees.

There were many pre-conference workshops offered, which I heard from other attendees were all excellent. I specifically attended the applied machine learning workshop taught by Max Kuhn, who developed the R package caret for machine learning. This workshop covered his recent work to improve upon caret in the recently released tidymodels R package, along with a set of related R packages (parnsip, tune, workflows).

Image

All the workshop materials are openly available on GitHub so anyone could learn a ton going through the PDFs in that repo! I'm looking forward to using what I learned in this workshop for analysis of time series data from the TERRA REF project and plant growth simulations for the Sentinel project.

The conference itself was also the best (and fanciest!) conference that I've ever been to. All of the presentations were extremely well put together, and there were plenty of opportunities to meet interesting, smart people working with R. All of the talks were recorded and are available on the RStudio website. Conference highlights included:

• A breakfast organized by RLadies that was so popular there was standing room only. There were interesting lightning talks on continuous integration, data science training, and art! I was one of the speakers (slides).
• An excellent talk by and conversation with Kate Hertweck, an educator at Fred Hutchinson Cancer Research Center, about community building and developing good training materials. I was able to pick her brain for some good low-hanging-fruit starting points for improving our group’s documentation, which can feel like an overwhelming task
• Emily Riederer's talk on incremental steps for taking Rmarkdown files from quick one-off use to better organized, more modular, and more reproducible products. See her corresponding blog post.
• I met some like-minded colleagues who also work at the University of Arizona. Kelsey E Gonzalez is a data science ambassador and social scientist while Carlos Scheidegger is a computer scientist interested in the TERRA REF data!
• RStudio announced that it has become a public benefit corporation to legally formalize its commitment to improving the R community.
• A presentation by Carl Howe about how data science training can be used to target issues of increasing amounts of data and scientific reproducibility over the next two years. 
• The livecode R package by Colin Rundel, which is used to provide live coding content to learners in a browser immediately while teaching.

This list only scratches the surface, as there were many, many more good presentations and chats at rstudio::conf. This conference did a great job of representing both the very supportive, engaging R community and cutting edge R development. If you missed it, check out the recorded talks from the conference!