Matrix computations (2nd edition) , by G. H. Golub and C. F. Van Loan. Pp 642. £38. 1989. ISBN 0-8018-3772-3 (John Hopkins Press)

A. Iserles
DOI: https://doi.org/10.2307/3619868
1990-10-01
Abstract:section briefly reports on the result of having only nine negative fundamental discriminants with class number 1, and gives tables for the remaining 56 fundamental (32 nonfundamental) discriminants of one class per genus. The chapter then ends with a discussion on the values of the Dirichlet's L-functions at the point 1, but this will only make sense to those readers who are familiar with the class number formula and the Generalised Riemann Hypothesis. The last often chapters is on factorisation and deals with various factorisation procedures related to the theory of binary quadratic forms, such as the SQUFOF method of D. Shanks. Throughout the book there is a reasonable number of concrete examples to illustrate the theory and there is also a fair number of tabular results; but it must be added that what the author considers to be algorithms are no more than proofs of theorems, if these are given at all. There are now many good word processors for authors to prepare "camera-ready" versions of their manuscripts. This is a new found freedom, but there is also an important responsibility associated with this freedom, and authors should seek proper advice and assistance from experts until they have mastered the skilful art of typesetting. The author here uses Tr^X, the most sophisticated computer typesetting program specially designed for mathematics printing, but he has left many glaring faults such as, to name the serious ones, the erratic "baselineskips" after each displayed formula, the distraction of an unacceptably large number of hyphenated words on many pages, and the lack of use of "ties" to prevent silly breaks.
Mathematics
What problem does this paper attempt to address?