It's a good book. $110 for the hard bound book is not so bad — I've paid over half as much for books that are not half as useful — but $105 for the electronic version is too much. Also, there is no discount for buying both PDF and paper.

I was very disappointed with this book. Save your money for something else. There is a good review of the problems with the book in Statistical Papers (2009) 50:445–446, reviewer: Uwe Ligges (http://www.springerlink.com/content/l36754377r182731/).

Dalgaard or Maindonald/Braun are much better options if you want a book about R.

For a total, starting-from-square-one newcomer to R, Crawley's book is the best. It's conversational in tone and gives many extended examples that cut through a lot of the R language confusion. For instance, Venables and Smith have two pages on data frames and then it's, "Good luck!" Crawley has almost 30 pages on them, with practical advice on how to get your data into a data frame.

I can see how advanced users will scoff at the book and there are some eccentricities, but for me it's been a lifesaver.

While I have not seen this book, have no idea who the author is, and would hardly ever pay more than 40 dollars for any programming book, I can't respect a review such as the one in Statistical Papers. Half of it is a not-so-interesting rant about the title of the book and the number of references (meanwhile, Stroustrup's 3rd edition C++ book had 2.5 pages of C++ references in its 912 pages…). The other half gives one (1) example of “bad programming practice'' for using T instead of TRUE, and concludes it can't recommend the book because it induces bad practice…

The only concrete thing I got from it is that the reviewer really hated the choice of title.

It's a good book. $110 for the hard bound book is not so bad — I've paid over half as much for books that are not half as useful — but $105 for the electronic version is too much. Also, there is no discount for buying both PDF and paper.

Is this (finally) a book that explains the basics of using R?

I was very disappointed with this book. Save your money for something else. There is a good review of the problems with the book in Statistical Papers (2009) 50:445–446, reviewer: Uwe Ligges (http://www.springerlink.com/content/l36754377r182731/).

Dalgaard or Maindonald/Braun are much better options if you want a book about R.

Wow–pretty brutal review! Perhaps in that 1000-page book there's a useful 100-page book of advice struggling to get out!

Tt is a bad book, Rnews criticized it.(page53, Volume 7/2, October 2007)

Well, that's Uwe for you. He is always direct. Fritz Leisch also reviewed the book in R News vol 7/2 (October 2007).

For a total, starting-from-square-one newcomer to R, Crawley's book is the best. It's conversational in tone and gives many extended examples that cut through a lot of the R language confusion. For instance, Venables and Smith have two pages on data frames and then it's, "Good luck!" Crawley has almost 30 pages on them, with practical advice on how to get your data into a data frame.

I can see how advanced users will scoff at the book and there are some eccentricities, but for me it's been a lifesaver.

While I have not seen this book, have no idea who the author is, and would hardly ever pay more than 40 dollars for any programming book, I can't respect a review such as the one in Statistical Papers. Half of it is a not-so-interesting rant about the title of the book and the number of references (meanwhile, Stroustrup's 3rd edition C++ book had 2.5 pages of C++ references in its 912 pages…). The other half gives one (1) example of “bad programming practice'' for using T instead of TRUE, and concludes it can't recommend the book because it induces bad practice…

The only concrete thing I got from it is that the reviewer really hated the choice of title.

Give Uwe a break, his own book has been roughly handled by a commenter on Amazon.