Mathematica Cookbook

Chapter 12. Statistics and Data Analysis

Watching in a trance The crew is certain Nothing left to chance ... Starting to collect Requested data “What will it affect When all is done?” Thinks Major Tom

—Peter Schilling, “Major Tom (Coming Home)”

12.0 Introduction

Ask statisticians what software they use, and chances are (no pun intended), they will mention SAS, SPSS, or maybe even R. Those systems are quite good, but most are highly specialized for statistical work. With the release of version 7, Wolfram has substantially beefed up the statistical capabilities of Mathematica. Given everything else Mathematica can do, it is now a compelling alternative for statistics and data analysis. An entire Mathematica statistical cookbook could be written; therefore, this chapter is necessarily incomplete. I have selected these recipes for this chapter to provide jumping-off points for further exploration. You should consult the Mathematica documentation for more depth, and nonexperts should consider Sarah Boslaugh and Paul Andrew Watters’ Statistics in a Nutshell (O’Reilly) for a broad overview of the relevant concepts.

Even readers without much interest in statistics are encouraged to skim these recipes because there are demonstrations here that have application outside statistics proper. Most users of Mathematica are comfortable with basic statistical metrics, such as mean and variance, but perhaps you are rusty on quantiles. All are covered in 12.1 Computing Common Statistical Metrics of Numerical ...