Product Description

This is the revised and expanded second edition of the hugely popular Numerical Recipes: the Art of Scientific Computing. The product of a unique collaboration among four leading scientists in academic research and industry, Numerical Recipes is a complete text and reference book on scientific computing. In a self-contained manner, it proceeds from mathematical and theoretical considerations to actual, practical computer routines. With over 100 new routines, bringing the total to well over 300, plus upgraded versions of many of the original routines, this new edition is the most practical, comprehensive handbook of scientific computing available today. The book retains the informal, easy-to-read style that made the first edition so popular, even while introducing some more advanced topics. It is an ideal textbook for scientists and engineers, and an indispensable reference for anyone who works in scientific computing. The second edition is available in FORTRAN, the quintessential language for numerical calculations, and in the increasingly popular C language. Not simply a line-by-line translation of the FORTRAN routines, the C version takes advantage of many unique features of the C language, including dynamic memory allocation, modularization, pointer reference to matrices, structured programming, and much more.

Product Details

• Amazon Sales Rank: #660671 in Books
• Published on: 1992-10-30
• Original language: English
• Number of items: 1
• Binding: Hardcover
• 994 pages

Review
‘If you are a programmer you need it. If you have someone to write programs for you, buy it as a birthday present.’ Spectroscopy Europe

Customer Reviews

The maths are OK but the code stinks
This book provides a broad coverage of numerical methods to solve a similarly broad spectrum of real world problems. The anlysis is not deep, but concentrates on matching problems to potential solutions. My main criticism of the mathematical side is that it seldom covers the criteria that make some numerical methods inappropriate in certain situations.

The major flaw in the book is the C code. It is woeful. The authors are, quite clearly, not C programmers at all. In fact, it looks like they are FORTRAN programmers who still cling to good old FORTRAN IV. When this FORTRAN coding style is allowed to be expressed in the freewheeling grammar of C the results prove to be monstrous.

I would recommend that C programmers buy Numerical Recipes in FORTRAN and do their own translations into C. Those transalations could not be any worse than the code in this book.

Useful reference but the coding is reprehenisble
Whilst the mathematics behind the book is undoubtedly useful, it is vitally important to understand the maths before using the methods presented in the text. An invaluable tool in the hands of the right people, and a weapon fatal to good science in the hands of the wrong people. Still, the text is suffiently advanced in its coverage to prevent most of the truely dangerous from using it.

Additionally, it is a shame that the authors/publishers did not see fit to have the code written by someone with real skill in this area. The code samples within the text are unfortunately typical of scientists who learned a little of programming techniques, probably in the 1970s and early 1980s, and have not really upgraded their understanding since. Whilst the code is functional and does the minimum required of it, it sends shivers down the spine of those more used to modern code practices.

Indispensable
Every chapter is a concise, readable and informative introduction to, and overview of, a field of numerical computing. If you need to use numerical methods, without being an expert on them, 'Numerical Recipes' acts as an unusually helpful tutor. The concepts are well explained in a way that makes clear the motivation, the strengths and the potential weaknesses of each method. It helps that the authors offer opinions and experience as well as mathematics. The structure of the book is good - it is easy to find the chapter you want, and easy to read each chapter, or section, without having to cross-reference other sections (I hate it when textbooks do that). I often read this book to choose my method, then use another subroutine library for implementation.