This work teaches the design of programs for scientific computation in C++. It introduces unique C++ classes, defines the particular relationships among these classes, and demonstrates their use in a dozen of the most powerful current applications. It presents a set of practices that allows programmers to embrace the attractive features of C++ without incurring undesired side effects and hidden costs. It includes a collection of source code files down loadable from the Wiley ftp site. It is originally announced as "Scientific Program Design: C++ for Native Fortran Writers".

Table of Contents

Preafce. Overview. Vector ad Matrix Basics. Sparse Matrix Basics. Conjugate-Gradient Methods. Triangular Matrices. Householder Matrix Methods. Singular Value Decomposition. Cholesky Decomposition. Automatic Derivatives. Constrained Optimization. Interior-Point Extensions. LU Factorization. Complex Arrays. Fourier Transforms. References. Index.

About the Author

JOHN R. BERRYHILL, PhD, earned his doctorate in physics from the University of Wisconsin-Madison. He has been writing programs and developing data-processing methods for over thirty years. He lives in Austin, Texas.

Reviews

"...teaches the design of programs for scientific computation in C++." (SciTech Book News, Vol. 25, No. 4, December 2001) "With a view to teaching the design of programs...the author introduces unique C++ classes for vectors, dense matrices, and sparse matrices...demonstrates their use..." (Mechanical Engineering, www.wemagazine.org, April 2, 2002)