Open Analytics @ UseR! 2015


The knarrs of Open Analytics have left the port of Antwerp on their way to Denmark. What will our delegation bring to Aalborg besides our loyal sponsorship?


For starters we’ve just released a Dockerfile Editor which may be particularly useful for Dirk Eddelbuettel’s tutorial on Docker on June 30. An overview of all tutorials can be found here.


On the first day of the conference Willem Ligtenberg will present at 16:00 on how to use databases in R without a line of SQL with Rango. The Databases session will see the first public appearance of an object-relational mapper (ORM) for R.

At the same time (16:00, Kaleidoscope 3) Tobias Verbeke will demonstrate our flagship product Architect, a full-fledged integrated development environment for data science. Keep an eye on our home page: rumours say a new release is approaching and that it is packed with new features..

For the more statistically inclined, Raphaƫl Coudret will present his implementation of the SAEM algorithm for left-censored data as part of the Lighting Talks.

Our final contribution for Wednesday will be a number of posters at the poster session starting at 18:00. As always, Jason Waddell has some insightful visualization experience to share and this time he will focus on density traces in plot legends. Maxim Nazarov has a poster on a suite of R packages for toxicological analyses following recently updated recommendations from the OECD. Meryam Krit, finally, presents goodness of fit techniques for exponential and two-parameter weibull distributions. She provides fast implementations for existing and novel methods that have, above all, been nicely packaged!


On Thursday Arunkumar Srinivasan kicks off at 10:30 (Kaleidoscope session) and will show how to push the limits of data manipulation with the recent features in the data.table package. It will be fast. It will be flexible and it will be memory-efficient!

Pushing the limits is also what Marvin Steijaert will talk about (10:30; Medicine session) be it on how predictive pipelines are deployed in pharmaceutical companies to push the frontiers in drug discovery using phenotypic deconvolution.

Kathy Mutambanengwe presents at 10:30 as well (Regression session) with innovative statistical work on spatial regression of quantiles based on parametric distributions.

Thursday at 13:00 (Visualisation 1) Kirsten Van Hoorde will demonstrate multiclass calibration tools in her new multiCalibration package. At 16:00 (Visualisation 2) Laure Cougnaud shows how she stays on top of high-dimensional genomic datasets using interactive versions of known and less well-known visualization techniques. Friday?

What about Friday? No more contributions from Open Analytics. Why? Conference dinner on Thursday evening :-)

Looking forward to a great conference!