The Twitterverse responds: Numerical Linear Algebra textbook recommendations.24 Dec 2021
I’ve become quite interested in numerical methods / scientific computing / applied math topics for a particular research question, and reading research papers in applied math wasn’t getting me where I wanted to go – and my graduate course in optimization covered algorithms, but not numerical issues.
So, I asked the Twitterverse for their recommendations, and was surprised at the volume of the response. Here’s the top selections, based on frequency and also whether it is a text; I learn better from books versus video lectures. The link goes to ISBNSearch or the book’s website. I like ISBN search because it lists vendors for the book in new and used, which for textbooks, is welcome.
- Lloyd N. Trefethen; David Bau III. 1997. Numerical Linear Algebra. ISBN 9780898713619.
- William H. Press, Saul A. Teukolsky, William T. Vetterling, Brian P. Flannery. 2007. Numerical Recipes: the Art of Scientific Computing. Website, with ebook. ISBN 9780521880688. (I am reading a 1988 version of this book while waiting on items to arrive.)
- Gene H. Golub and Charles F Van Loan. 2013. Matrix Computations. ISBN 9781421407944.
- Jorge Nocedal and Stephen Wright. 2006. Numerical Optimization. ISBN 9780387303031.
- James W. Demmel. 1997. Applied Numerical Linear Algebra. ISBN 9780898713893.
- Dianne O’Leary. 2009. Scientific Computing with Case Studies. Website. ISBN 9780898716665.
- Yousef Saad’s works: Iterative Methods for Sparse Linear Systems. 1996. ISBN 9780534947767 Numerical Methods for Large Eigenvalue Problems, 2011. ISBN 9781611970722. See the website for updates and ways to access these works, as they are both out of print.
- J. Stoer and R. Bulirsch. Introduction to Numerical Analysis. 2010. ISBN 9781441930064
An observation: many of these books are out of print and the prices, even used, are very high. Others are provided free by the authors, so long as you don’t mind reading from a screen and/or contract to get the book printed and bound.