Statistics with Mathematica
Martha L. Abell, James P. Braselton, and John A. Rafter, 1999, Academic Press. 632pp, softcover, w/CD-ROM.
Statistics with Mathematica fills what has been a somewhat large gap in the range of applied Mathematica publications, that of statistics. While the Mathematica Standard Add-on Packages provide a helpful set of tools in this regard, this book helps tie together several aspects of Mathematica and demonstrates the program's often-neglected usefulness for performing statistical analysis. Starting with the obligatory Mathematica tutorial material (including a tear-out Quick Reference card), it walks the reader through data manipulation, univariate and multivariate descriptive statistics, probability, ANOVA, regression and correlation, diagnostics, time series, simulation, and non-parametrics using real examples.
The overall design of the book is readable but it suffers from the all-too-common problem of having to integrate lots of standard Mathematica InputForm and OutputForm cells, graphics, tables and ruled areas with the body and heading text and equations. The PC-centric CD-ROM (readable on the Macintosh) contains the data sets and Mathematica code used in the text. Of interest to many readers will be the fact that these notebooks include some functionality left out of the Standard Add-on Statistics Packages such as non-parametrics and multiple comparisons. Unfortunately they are not in standard package form, so you must wade through the code a bit in order to find what you want and shoe-horn it into your analyses.
While it provides a moderate amount of didactic material on statistics, the book's main strength is its illustration of how to perform statistical calculations using Mathematica. Therefore, it would make an excellent supplementary text for a statistics course that uses Mathematica. The material covers a rather broad range of topics and thus should have an equally broad appeal to a variety of levels of statistical understanding. It would probably not make a very complete textbook on its own (and, to be fair, the authors never claim that it would), but if you are already familiar with statistics at almost any level you'll find this book a useful companion in your analytical endeavors.
-- Flip Phillips
Contents copyright 1999