R Statistics Package

R refers to the statistical package developed by the R Project for Statistical Computing1. It has gained traction recently as a viable alternative to many of the most popular statistics packages such as SPSS, SAS, and S+2. This popularity is strongly related to a few factors, namely that is is free and open source, available for Windows, OSX and Linux, and that it is extremely robust and extensible. Because of the free nature of R a new generation of academic researchers and grad students are learning and developing on it, whom are now filtering into industry. R has built in capabilities to perform linear models, graphics and many other statistical needs. Because R is a language, it is relatively easy to write custom functions in R to tailor an analysis to a specific type of data, automate frequently run tasks or to develop a brand new analysis.

Specific Sensory Science Add-ons

There are hundreds of add-on packages available for R which provide functionality ranging from multivariate statistics to time series analysis, genetic algorithms and neural networks. The Comprehensive R Archive Network (CRAN)3 provides a thorough database for each package. In the past few years, some packages have been developed which are specifically designed to aid the sensory practitioner. These include:

Package Name  Functionality
FactoMineR4  Exploratory data analysis and multivariate statistics, including MFA,  Generalized Procrustes Analysis (GPA) and PCA.
SensoMineR5 Sensory DOE, Projective Mapping analysis, and other sensory functions
SensR6   Thurstonian models for sensory discrimination.


Other packages

The RCmdr7 package provides a rudimentary GUI interface to basic R statistical functions including basic linear models. It also aids in the importing and exporting of data to and from .xls or .csv files.

ggplot28 provides extremely advanced graphics capabilities. Sweave9 provides integration of R code into LaTeX (a typesetting language) documents, and in the newer versions of R is built-in. This allows the creation of high quality dynamic statistical reports on-the-fly from raw data, vastly increasing the efficiency of running and reporting routine sensory data.

References

 
 
2 Vance, Ashlee. Data Analysis Captivated by R’s Power. The New York Times, 2009-01-06
 
 
4 FactoMineR Home Page
 
5 SensoMineR Home Page
 
6 SensR Package Page at CRAN
 
7 RCmdr Package Page at CRAN
 
8 ggplot2 Home Page
 
9 Friedrich Leisch. Sweave: Dynamic generation of statistical reports using literate data analysis. In Wolfgang Hardle and Bernd Ranz, editors, Compstat 2002 – Proceedings in Computational Statistics, pages 575-580. Physica Verlag, Heidelberg, 2002. ISBN 3-7908-1517-9.