Numerical algorithms with Fortran /

Numerical Algorithms is not a textbook on numerical analysis: there are many theorems, but no proofs; and many algorithms but very few examples and no problems. This is a resource for mathematicians and scientists needing solutions in a great variety of mathematical areas, some of which can be found only in the primary literature. Principles of the various methods and descriptions of the algorithms are included. A CD-ROM containing Fortran programs of these solutions is also included; viewers will need Windows and a World Wide Web browser to use it. The following topics are included in the 18 chapters: computer representation of numbers, error and conditioning analysis, stability of algorithms; methods to solve nonlinear and algebraic equations; direct and iterative algorithms for solving linear systems of equations; systems of nonlinear equations; eigenvalues and eigenvectors of matrices; linear and nonlinear approximation; polynomial and rational approximation; interpolating polynomial splines and cubic fitting splines for generating smooth curves; two-dimensional splines, surface splines, Bezier splines, and B-splines; Akima and Renner subsplines; numerical differentiation, integration, and cubature; and initial and boundary value problems for ordinary differential equations. Professional. H. J. Bender University of Maryland University College