The information below will be periodically updated at the folowing permanent link: http://www.backsidesmack.com/r-resources/ Searching for information on R sucks. Not only is the language name a letter of the alphabet (an ignominy it shares with C and some less well known languages), there is Pearson’s r and the coefficient of determination, R squared! if you don’t believe me google “R stats covariance” (or some similar frequently asked question) and see. So I figured I would provide some helpful resources for someone learning R:
- CRAN Task views: Available here, they are a somewhat curated collection of R packages by use. Not much help if you are totally rudderless but once you have an inklink of how you want to tackle a problem they can point you to the right tools.
- R Bloggers: If you work with R and you have an RSS reader this blog should be on your feeds. R-bloggers aggregates feeds from across disciplines and skill levels so one day you can read about someone like me struggling with matching country pairs and another day you can read about using R to support research on climate change. Less a single resource than a continuous source of tools and tricks.
- CRAN contributed docs: This collection of documentation on R is a bit harder to recommend because it contains so much great information. So long as you focus on guides which suit your discipline you will do well. There is some duplication between guides (especially with the earlier stuff like getting R running or setting a library path), but most of them are excellent and all of them are free.
- THE R INFERNO!: Patrick Burns’ awesome guide to all the quirky things in R that can screw you. Some of the material might be familiar to readers (namely the FP vs. integer issue), but most of it is the sort of received wisdom we wish were written down. Newly updated for 2011 this resource covers (almost) every darkened corner of base R; containing sage advice for migrants from C style languages, a whole chapter on when not to write functions and the problems that can crop up in testing all in a clever presentation.
- Books: I’m hesitant to recommend books on R for a variety of reasons, namely that most good introductions to R come packaged with an introduction to a specific discipline (or an introduction to statistics). This includes Jay Kerns’ excellent and f/Free Introduction to Probability and Statistics Using R (downloadable as a PDF or you can get the package and build your own with Sweave!). If you want to learn statistics and R simultaneously, great. There is no shortage of excellent texts out there for you. If you are looking for a resource on the language you may find the R Cookbook by Paul Teetor useful. I haven’t read it but I have heard nothing but good things. I love John Chambers’ Software for Data Analysis. It is clear, interesting and in depth. It is not a desk reference, so be warned.
- Q and A websites: Both Cross Validated and the statistics sub-reddit are good places to ask questions without too much fear that someone will bite your head off. In exchange for ease of use you give up the chance to have your question answered by Peter Daalgard (as you might on r-help), but they are interesting enough.
- Statistics with R: Now here’s one you probably haven’t seen. Statistics with R by Vincent Zoonekynd is a sprawling mess of a website complete with giant title image and a yellow background for quoted text. And it is amazing. If you want a step by step look at regression diagnostics complete with graphs and detailed code examples, this is your website.