Dear MilanoR friends,
I am just back from UseR 2013 in Albacete (Spain) and I would like to share.
First of all I would like to thank all staff from the conference organizer committee and personnel for the excellent job they have done in organizing and running the conference.
Secondly, a special thanks goes the the conference sponsors for their generosity that made conference time always a special and tasteful time.
The conference was made of a large variety of interesting talks and, choosing between them has been a very hard job.
Because of my personal interests, I privileged talks on computational statistics and programming techniques rather that applied statistics and, among the talks attended I would like to point out few of them:
Hadley Wickham - BigR data
R and "bigger" data; how does R deal with big data? Well, according to Hadley's speech, R performs quite well.
The whole presentation can be found here.
Duncan Murdoch's - What's New in R 3.0.x
A clear and informative presentation on the newest R 3.0.x and its significant improvements from previous 2.x.x version. Duncan also provided some insights about R 3.1 future release regarding its integration with "
pqR" package from Radford Neal.
I am not sure where to download the original presentation that was given in Albacete but, a very similar presentation can be found here.
Rasmus Baath - An expose' of naming conventions in R
An interesting piece of research within R and its variety of naming conventions, showing how many naming conventions can coexists within the same system.
Colin Gillespie - Reference classes: a case study with the
Reference classes in R allows mutable objects, and are particularly useful when dialing with complex simulation and GUI development
My personal award along with my sincere gratitude to the Rstudio guys for the implementation of a visual debugger integrated within Rstudio and Rstudio server. Surely, a great help to the whole R community.
Finally, I would recommended to pay particular attention to Shiny, a web application for deploying R to the web developed by Joe Chang again from RStudio.
In particular, the yet to be released "reactivity" package will surely make the difference when developing web based analytics with R.
Reactivity in fact, will allow to change the value of an existing R variable and have and have expressions, tables and charts based on that variable update instantly
A look to the future
I would conclude saying that I would be really pleased to see a future edition of useR conference to be organized in Italy for the Italian R community being so active and numerous.