After a successful EARL London in September of this year (see here and here) it was Boston’s turn to shine. This is only the second time this conference has run, yet its quality of speakers is at par with London. This year’s edition welcomed twenty presenters from eight industries and saw almost a hundred attendees, yet another showcase for R’s widespread adoption across all industries.
One of our fantastic keynotes at EARL Boston this year was Jenny Bryan, who has recently joined the RStudio team. Jenny used a series of visual analogies, which cleverly (and comically) combined Lego structures with Game of Thrones references to simplify data wrangling techniques. She also demonstrated to us the ‘map’ function from the purr package. She successfully managed to convert me from an ‘apply’ user, armed with nothing more than Lego men, with and without hair! In the case of more complex structures such as nested data frames, taking that extra step to stop and think about the physical structure of your data (Lego not compulsory) can make analysis seem instantly less daunting.
The next keynote speaker was Ricardo Bion from AirBnB who elaborated on how they use R to create reproducible research. Their way of working is quite impressive and now that they’ve open sourced their knowledge repo it is available to us all. Ricardo also mentioned they are using RStudio Connect and it so happened to be that Tareef Kawaf presented about RStudio Connect as well. This promises to be a very useful addition to the RStudio technology stack and one to watch for in the coming year.
After lunch Daniel Hadley showed us how far data scientists are willing to go for their data. He gave an anecdote about rummaging through bins in order to gather data on the amount of wastage per household with a view to simulating (in R) the required bin size for households across Boston. On a more serious note, Daniel’s talk really hit the theme of EARL – Effective Applications of the R Language – by discussing different packages and different use cases of the language across differently themed projects. Jared Lander and Tanya Cashorali further emphasized this point in their talks. It is great to see how R is used to make a difference in real world applications and decision making that affects considerably large parts of the population.
Now that Microsoft has stepped up its endorsement of R we can expect more and more applications of R with Microsoft products. We anticipated this and organised a successful workshop around it. David Smith also enlightened us with an overview presentation of what Microsoft has to offer and Danielle Dean, Jaya Mathew and Francesca Lazzeri showcased actual applications of R with SQL Server and Azure respectively.
Despite the involvement of some big companies we shouldn’t lose track of what R is for and where its roots really lie. For that, we only have to look at the presentations by Amar Dhand and Aedin Culhane. Amar showed how R can be used to apply network analysis on hospital networks and showed that healthcare costs increase as patients go from hospital to hospital. Aedin explained how R is used to support cancer research at the Dana Farber Cancer Institute. In dealing with big data in a complex environment R held its own. These are only two examples of what we try to achieve at EARL every year: to showcase truly effective applications of a dynamic language.