The revolution in computing has revolutionized statistics. I can still remember my nights at the computer center, running SPSS, SAS and BMDP on the huge mainframes. Many of us preferred to run the larger jobs at night when the mainframes were less busy: if you were lucky, you could have your job run with a huge memory slice, like 16K or 32K! Much faster, of course, than when you submitted jobs during daytime.
Then later, I was elated when I could get SAS and SPSS on my desktop. It was expensive but worth every penny, in my opinion. After all, now I could clean data, recode, and run my analyses whenever I wanted, without running to the computer center.
And now, with internet and powerful computers everywhere, we have R. And as professor Crawley says in this interview, I am very enthusiastic about it: R is here to stay! R is one of the many successes of open source – an incredibly powerful statistical software package that is excellently supported, extensively used, available for Windows, Mac and Linux, and completely free! Go to http://www.r-project.org/, find the version for your computer, download and get going.
The R language is today acknowledged as one of the most powerful and flexible statistical software packages, enabling users to apply a variety of sophisticated statistical techniques. The system has grown tremendously over a decade and a half, is huge, and full of smart and useful options. While it is possible to just start using it and googling when you’re stuck, I don’t recommend it. I know we are all different, but at least I – and I think many others – have found that having some basic navigational tools available speeds up the learning. So, just as I would use navigational maps on a new stretch of ocean, I use The R Book when working with R. It makes the learning smoother.
The R Book is a 1050 page book, very concise, well organized and with a good index. It first covers all the basics of the R language – data input and manipulation, graphics, tables, tests and statistical modeling. Then it delves into the more advanced subjects: basic and advanced methods for analysis of discrete and continuous outcomes, linear and non-linear analyses, as well as analyses of time and spatial dependences. Survival analysis and simulation modeling are covered as well. At the end of the book, Crawley demonstrates how to use R for graphics. While this part is much better than in the previous edition, it is still perhaps the weakest part of the book.
Overall, The R Book (Second Edition) is a great guide to the vastly powerful and constantly evolving software that is R. It is very close to a complete reference-the coverage is excellent. For most users of R, having this book as guide will make life with R much easier, and learning to master it much faster.