On the Edge: Statistics & Computing
The focus is on broadening the statistical computing from the traditional methodology and algorithmic focus to include data acquisition and input and the distribution of results in different ways such as "live" documents, Web-based tools, and so forth.
Abstract
As I write this from just south of Napa Valley, California's premier wine-producing area, I am drawn toward thinking of this article as a tasting tour. On the menu are short sips of the Omega projectand the associated software. I hope that at the end you will have a good idea about the different flavors and which ones appeal to you, and that you don't spit out your sample. Asis customary on such tours, we start with the docent's short speech on what we are about, our history, and what you can expect to encounter. Omega is an attempt to develop the next generation of statistical computing tools. The focus is on broadening the statistical computing from the traditional methodology and algorithmic focus to include data acquisition and input and the distribution of results in different ways such as "live" documents, Web-based tools, and so forth. The idea is to think of the statistical environment as a small network of interacting tools on the desktop or Internet that are used to acquire data, perform statistical tasks, distribute results, and so forth. This is in contrast to most existing systems, which are single monolithic applications. Additionally, the tools are designed to allow easy integration of other technology as it becomes available, to be embedded in other applications, and to communicate between applications. And finally, the design is for a flexible, extensible system in which a user can substitute for different elements with others more appropriate for her/his computations (e.g., algorithms specific to large datasets,