How can I get to speed with R for processing statistical data?

What is it?

R is a computer language and environment for statistical computing and graphics. It is widely used in biostatistics, including microarray and proteomics data analysis.

R provides a wide variety of statistical tests:

• linear and nonlinear modelling
• classical statistical tests
• time-series analysis
• classification
• clustering

R also provides graphical techniques, although native R is not strong on graphics [example 1] [example 2], and is highly extensible, meaning you write you can modify existing code or call complete new functions.

However, R is not for everyone. If you have only basic statistical needs, or are not at least somewhat familiar with programming, R is not for you. Perhaps a program such as GraphPad's InStat or Excel's statistical functions might be preferable.

S+, a commercial equivalent of R, is by far the better solution if you require significant graphing capabilities; this language is essentially identical to R and requires purchase of a moderately priced license.

What is it for?

Here are selected examples applicable to life sciences; details are available here:

• Microarray data analysis, particularly using BioConductor
• Bayesian Inference
• Cluster Analysis & Finite Mixture Models
• Analysis of ecological and environmental data
• Statistical Genetics
• Machine Learning & Statistical Learning
• Multivariate Statistics
• Spatial Analysis of Spatial Data
• Graphical models in R

Obtaining R

R is free and can be obtained here. It compiles and runs on:

• UNIX
• FreeBSD
• Linux
• Windows
• MacOS [example]

IMPORTANT: If you don't know what "compiling" means, R may not be suited to you:

• If you know some programming, you may want to want to consult Appendix E of Using R for Introductory Statistics (Verzani 2004), available from Lane. It will provide you with the essentials of the language in very concise form.
• If you don't know at least some programming, the learning curve for R will likely be very significant, and it might be preferable to use another tool until you are comfortable with at least one programming language.

You can also use R on computers in the M202 Computer Laboratory: all PCs in the M202 Teaching Lab have R already installed, usable by anyone with a SUNetID.

R training

1. Hung Chen's succinct overview of R programming
2. David Metz and Brad Hunting's excellent R tutorial:
   • Part 1: Dabbling with a wealth of statistical facilities
   • Part 2: Functional programming and data exploration
   
   Other selected training documents:
   • Introduction to R
   • R Data Import and Export
   • Installing R
   • Complete list of manuals

Key references

• BOOK: Verzani, J. (2004) Using R for Introductory Statistics .
• BOOK: Dalgaard, P. (2002) Introductory Statistics with R , Springer; good, brief overview of R's features applied to actual examples.
• eBook/BOOK: Bioinformatics and computational biology solutions using R and Bioconductor ,Gentleman R, (2005), available at the Lane Library.
• BOOK: Spector, P (1994) An introduction to S and S-plus ; excellent overview of the S language, essentially identical to R. Available from the Mathematical & Computer Sciences Library.
• Comprehensive R Archive Network:source code and executables.
• R home site.
• R FAQs.
• Search R documentation.
• All of Dr. Balise's videorecorded statistics classes
• Dr. Balise's top statistics books.

Source

Lane Librarian

Record created 9/21/2006; updated 1/12/2007.

ypouliot, September 16, 2009