The R Cookbook should be on your bookshelf if you work with R.
The book is as self-described, a collection of tasks and how to accomplish those tasks in R (recipes). This is not a tutorial on the language, but is definitely recommended for novices. One of the most frustrating aspects of R for the beginner is to know what manipulations you require for a dataset, but to be clueless as to how to perform those steps in R; this book can help close that gap.
For intermediate users, it can serve as a reference. I'll often use this to jog my memory as to how a particular technique is applied, e.g., run a function on each row of a dataframe. Since the book has been available on the O'Reilly Safari system for several months, it's become one of my most-used options for R info.
Technically the book appears to be accurate, with the recipes I've used functioning well. Caveat, I have not tested any of the higher-end statistical recipes, as they aren't required in my work.
In summary, this should be one of the first books purchased when building an R library.
Disclaimer, I received access from O'Reilly Publishing to an electronic copy of this book for purposes of review.