Textbook in PDF format

This collection of algorithms from the Royal Statistical Society's journal Applied Statistics is the first of its kind: it promises to become a standard reference work. Over 200 algorithms have appeared in the Algorithms Section of Applied Statistics over the last 16 years, and the time is right for this anthology which presents 32 of the best of them (together with one from another journal). The book bears the hallmark of quality that distinguishes the Royal Statistical Society's journals. All the algorithms presented have been thoroughly reexamined by the editors, taking into account all corrections and modifications that have been found necessary or desirable over the years. They have also been updated to follow the latest version of the journal's guidelines for the construction and description of algorithms. They can be used in "black box" style, ignoring the inner workings while presenting them with input and receiving reliable output, but those who write computer programs may also wish to study their detailed construction. Each algorithm is presented as a self-contained procedure, usually in Fortran strictly following the 1966 standard as well as the present definition — Fortran 77. Thus any Fortran compiler that follows these standards should be able to run them without modification. A few are in Algol 60, following the current standard for this language, but these are accompanied by Fortran translations to enable them to be used in conjunction with the others. An introduction to each algorithm gives details of its purpose, method of use and mathematical background. Readership: Mathematicians, statisticians, and all concerned with the relationship between mathematics and computers. Computer scientists, economists, management scientists and, as supplementary reading for students of courses in all these subject areas