Fri Oct 4, 2002

A Great Piece of Free Software

As I write, the latest version of R is compiling in the background on my workstation. R is a high-quality language and environment for statistical computing. It’s related to the S language developed at Bell Labs which is now marketed as the commercial package S-Plus. R is a terrific statistics package. I use it for all of my quantitative data analysis. It’s powerful, flexible and full-featured. It’s developed by a top-notch community of statisticians who provide amazing on-line support to users. It’s got a growing, and very strong, literature of introductory texts, companions to existing textbooks, implementing modern statistical methods, regression modeling strategies, specialized types of models and programming techniques. Its structure encourages the user to learn more about good statistical practice rather than relying on canned menu-driven routines. It’s available for a wide variety of platforms, including Windows, Macintosh, Linux and other Unix-based OSes and processors. And it’s Free. In short, it’s the best kind of open-source software project. You don’t hear about it very often because many Linux users care more about their desktop or their mp3 player.

I run R inside of Emacs, the powerful text editor. A plug-in for Emacs called ESS (Emacs Speaks Statistics) supports R and makes working in R very easy. (It does the same for S, SAS, Stata and other programs, too.) Here’s an obligatory