This book grew out of notes for my Statistical Computing course that I have been teaching for the past 20 years at North Carolina State University. The goal of this course is to prepare doctoral students with the computing tools needed for statistical research, and I have augmented this core with related topics that through the years I have found useful for colleagues and graduate students. As a result, this book covers a wide range of computational issues, from arithmetic, numerical linear algebra, and approximation, which are typical numerical analysis topics, to optimization and nonlinear regression, to random number generation, and finally to fast algorithms. I have emphasized numerical techniques, but restricted the scope to those regularly employed in the field of statistics, and dropped some traditional numerical analysis topics such as differential equations. Many of the exercises in this book arose from questions posed to me by colleagues and students.

Most of the students that I have taught come with a graduate level understanding of statistics, no experience in numerical analysis, and little skill in a programming language. Consequently, I cover only about half of this material in a one-semester course. For those with a background in numerical analysis, a basic understanding of two statistical topics, regression and maximum likelihood, would be necessary.

I would advise any instructor of statistical computing not to shortchange the fundamental topic of arithmetic.