I have always been fascinated by statistics, going back to the time at a summer Boy Scout camp when I was so bored with the physical activities that I started reading a little booklet, which I found just removed from the library. It was published in 1923, written in Polish with a title that in direct translation sounds something like Introduction to probability calculus and mathematical statistics [1]. The booklet was authored by Jerzy Splawa-Neyman, later a famous professor of statistics at the University of California, Berkeley. It was apparently an early predecessor of Neyman’s later classical book on probability and statistics [2]. This little booklet of around 100 pages explained to me the basic concepts of probability with examples of tossing a coin, pulling balls out of an urn, and picking an ace out of a card deck; it sparked my interest in statistics, which continued for some time into my college years, until it was later overcome by fascination with computer programming. And here is a book on R, which merges both!

Having heard about the R programming language, I initially wondered, why introduce a new language, with a focus on statistics, when so many statistical tools exist already? There are complete packages, such as SAS and SPSS, and mathematical tools, such as MATLAB, Mathematica, and Maple, which include an array of statistical features. The answer came to me quickly when I started using the language: it is a free alternative to expensive statistical software. It has been gaining popularity since its creation in New Zealand, in the mid-1990s, as a clone of Bell Lab’s S language. In fact, it is not just a language but an entire environment, with expanding functionality available in libraries and packages, with new features being constantly added to its already rich function set by the user community.

The book itself appears to be a very important part of this environment because it is a complete educational tool on learning and using the language and applying it to statistical problems. Looking from the perspective of a novice reader, one should first install the package, following instructions given in the first chapter and in the appendices. Then, the reader should be ready to learn the language itself.

The constructs of the language are presented in the first two parts of the book. They cover, respectively, data structures in R and language operations, all illustrated with numerous examples. The next two parts deal with introducing the concepts of statistics and probability, and applying the language to statistical modeling, that is, hypothesis testing, analysis of variance, linear regression, and diagnostics. The purpose of all chapters in this part of the book is to explain how to use the language in solving statistical problems and show the benefits of applying it as a tool. While graphical presentation is an intrinsic part of presenting the language as well as applying it, the fifth part of the book is focused exclusively on more advanced graphics.

This book contains over 750 pages of text, so it cannot be read cover to cover in one sitting. But because it has a clearly defined structure, one can easily focus on aspects of specific interest, which could be, for example, first learning the language, and then progressing with its application to address selected problems in statistics. And this is, in fact, what the author intended: to serve a variety of audiences that would be interested in the dual aspects of using R as both a programming language and as a tool for statistical problem solving. In this regard, the book serves these audiences well, including students, researchers, and practitioners of both computing and statistical methods.

More reviews about this item: Amazon