This book describes itself as a ‘cookbook’ for non-linear regression and is supported by the nls2 software that has been made available on the Web site (http://www.inra.fr/bia/J/AB/nls2/) of one of the authors. It is very much in the style of a computer manual and features S-PLUS or R code in a Courier fount which can be copied and executed. This extends to the requirement to enter the data sets manually. The chapters are reasonably and logically laid out, but it must be stressed that it is unashamedly a cookery book. There are more theoretical sections sprinkled around the book which have been differentiated by a sans serif fount. Thus some issues (such as bootstrapping) are given as practical examples but the rationale is left to be followed up by means of external references. There is a short section 4.4 (pages 114–115) discussing numerical problems which may arise during the estimation process: where alternative models must be used the reader is referred to Ratkowsky (1989).

The book uses alphanumeric citations, such as ‘GJ94’ rather than the more conventional ‘Gruet and Jolivet (1994)’. There are 42 references, many of which are to other text-books on modelling; the indexing is sparse. The back cover suggests that it may be of use to students as a tutorial book. It is certainly a valuable complement to the nls2 software: in the second edition this includes the ability to fit binomial, Poisson and multinomial models. The software installed without any problems on Debian Linux working with R, but it should be noted that it is not available via CRAN. There are also S-PLUS and standalone versions of the software.

Reference

Ratkowsky
,
D. A.
(
1989
)
Handbook of Nonlinear Re-gression Models
.
New York
:
Dekker
.

This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)