The Technical Corner

by Steve Craighead

This is an inaugural column in Expanding Horizons. Our intent is to inform you, our readers, to available open source modeling software.

Our first series will revolve around the use of the R language . R has become the lingua franca of statistical and data modeling. It is the outgrowth of the original S language and its commercial cousin S–Plus. R is maintained by a cadre of volunteers constantly testing and improving its modeling capabilities. One of the benefits of R is that as researchers develop and publish their new statistical and modeling algorithms, frequently these are implemented within the R language.

R is designed within an object oriented designed framework, and is the ultimate in flexibility and code reuse and extensibility. R also comes in your favorite choice of operating system, such as Windows, Linux and MAC.

In an R worksheet one first collects or imports their data. A description of the flexibility of data importation or exportation could take up several future columns alone. After the data is imported, there is a plethora of different analytic and graphical tools available within the base library alone. At this point in time, the official current release (2.4) of R has 25 base libraries and approximately 950 contributed libraries. These libraries cover a wide range of computational models of which Generalized Linear and Additive Models, Time Series, Independent Component Analysis, Quantile Regression, Option Pricing, and Portfolio analysis are just a few. In fact Vincent Goulet and S�bastien Auclairone are the authors of the actuar R library package. Possibly, either one of them could write a future column on their work.

In our next column, we will examine the various Option Pricing libraries and observe their usefulness.

If you want to start looking at this language on your own, go to either R–Project.org or Cran.R–Project.org Web sites. There is extensive documentation and you can download your own working copy!