This book is an outstanding introduction to practical numerical methods for (budding) physicists who have no experience with these vital tools. The author states the book is aimed at undergraduate seniors or first-year graduates. This seems pessimistic to me: I think any competent undergraduate who has taken a course in ordinary differential equations could hack it.
The book ignores the usual approach taken by numerical analysis texts, which is to build up from the fundamental ideas (e.g., finite precision arithmetic, error propagation, fixed point iteration, finite difference approximation to the derivative), instead jumping almost immediately into a projectile motion ODE problem. This allows the author to move on quickly to adaptive Runge-Kutta in Chapter 3, Fast Fourier Transforms in Chapter 5, PDEs in Chapter 6 and finish with a discussion of Monte Carlo methods; whereas more traditional books will only begin to cover PDEs near the end and usually do not discuss FFTs or Monte Carlo.
Of course, this comes at a price. I took a senior level course taught in the traditional manner described above, and happened to pick up a copy of this book in the middle of the semester. This book has far more physical insight than my assigned text, and leaves the student able to appoach a far greater set of practical problems, but I think those who are serious about computational work should cover the basics more thoroughly. One outstanding feature of the book is the end of chapter projects that unify and apply what has been learned, and offer a chance for better students to stretch their muscles.
On the other side, what the author says in the preface bears repeating here: the methods in described in this book are (almost all) foundational, and nowhere near the state of the art. This is particularly true of the relaxational methods for PDEs described in Chapter 8. Nor do I think this would make a very useful reference book: anyone experienced enough to be able to read and understand (say) Numerical Recipes will not learn much from this book. Also, for a modestly-sized paperback with only black-and-white printing, it is amazingly expensive.