Numerical Recipes 3rd Edition: The Art of Scientific Computing

I own the C version of this book so I knew it was going to be of the highest quality but I was still caught off guard. One improvement over the C version is the fact that this book has no dust cover. The title and other information is printed directly on the cover and that makes me less likely to drop the book. Books this large tend to slip out of their dust covers. I have not read this book yet but I flipped through it and noticed that some of the math is rather advanced. That's OK because I love math! This looks like more of a math book than a programming book like those written by Knuth. I made a study plan that lists the books I'm going to read in a thought-out order. My plan does start with C++ but it will still take a couple of months before I get to this book. After I read it, I may come back here and update this review. I wish that I could give it a 6th start already and I haven't even read the book yet.

I have owned copies of the first two editions of this book and I was impressed with the updates to this third edition. There are several new topics and the existing areas have been updated. I was able to use some of the statistical code in a production piece of software two days after receiving this new version. The website associated with the book also has a nice feature that figures out the header dependencies for you.

I've bought several editions of this book: the original one (Fortran code), the C version, and now this latest in C++. I bought this last version mainly for new material on the algorithms, and not for the C++ code.The book is mainly useful for its discussion of the issues involved in numerical analysis, the presentation of algorithms, and -- to some extent -- the demonstration of the algorithms in code.That said, the biggest weakness of this book is the code itself. The C++ programs appear to be direct translations of Fortran, bringing to mind the old joke that 'real programmers can write Fortran in any language'. The book would have been far more useful, and the implementations far easier to understand, if the algorithms had been simply written in pseudocode. The code is insufficiently commented, and the variable names are often hard to comprehend. In fact, the code in this book violates almost every precept laid down by Steve McConnell in his excellent -- and still relevant -- book Code Complete 2nd ed.In that regard, the authors' concerns about potential copyright violations of their code are almost laughable; no sane programmer writing in C++ (or in Java or C#, or any other modern language, for that matter) would write his/her code the same way as presented in the book.I give this a 4* rating because of how comprehensive this book is and the care the authors take to discuss implementation issues. But I almost knocked it down to 3* because of the horrible coding.

I read parts of this book for knowledge of model building and matrix inversion. I learned a lot by reading the text on these subjects, which I had never done for editions 1 and 2. The paper is very thin; it is easy to deface a page by spilling a liquid on it. Also pages tear easily. The examples actually work, which, as we know, is not true of all computer science books.

Anyone looking at this review already knows what this book is. It's an updated version of the code. Personally, I find C not helpful. Fortran code is more useful for me. I translated that code into Go.

This book is the greatest compilation of computational knowledge that any seasoned or unseasoned computer scientist, and student, could ever have. I have been programming for 15 years and I still go back to this book as the first step in learning or re-learning any computational technique. I own this book in both C and C++, because it's just that amazing.

This or a similar book is a must if you do mathematical programming. Algorithms are stable and very fast. That’s on the positive side.On the negative side,1) the physical book’s quality is not great, the binding easily breaks2) the code, while efficient, is barely readable. It is old Fortran style with variables aaa, bbb etc. So you can copy the code, but will have hard time altering it for your needs.3) on occasion the book does not contain the whole algorithm and prompts you to go online.Well, I could go online without paying $100 for the book, and expected the book to be sufficient.Having said this all, I’m still keeping it.

I never bother to use the code from this book, but I've been using the second edition for years as a very solid mathematical and numerical reference. For me, this book is a critical reference and usually the first place I turn to when I need additional insight into methods I need but am not sufficiently familiar with.