Wednesday, January 07, 2009

R: A not so silent revolution

Many people don't know what R is. However, things are changing fast for a simple open source software that has long been known to graduate students in technical programs and PhD students at many universities. R is an open source statistical software program originally developed by academic researchers. It is robust, flexible and more importantly it allows anyone interested in statistical analysis to have access to the latest in computing solutions through a simple user interface. The beauty of R is that it comes packaged with the community's knowledge. Apart from the meticulously maintained email archives, forums and FAQs a user can download and use thousands of "packages" or routines that have been written by the community.

There is a package for every possible statistical scenario that even a sophisticated user might need. If the problem is too specific then one can either program the solution from scratch or use one of the existing packages as the starting point to build a new solution. The beauty of open source allows for these modifications. Additionally, majority of the packages come with detailed user manuals as well. This article on NYT reports on the ubiquitous presence of R in organizations such as Google. A majority of technically savvy engineers and programmers have been trained on R. Graduate course work in many universities uses R as a programming tool for every conceivable task. This has helped spawn a new generation of programmers who are beginning to use R as a tool at firm like Google which needs to sift through trillions of bits of data. Sophisticated models can then be used to analyze such data for generating BI.

Increasingly, firms will face an overload of information and merely collecting and storing data cannot bestow competitive advantages. Large firms are at an advantage because they will have the resources to buy and use proprietary software tools that will allow them to analyze massive amounts of information and hopefully generate BI. However, there will be millions of small and medium sized firms that will not have such a luxury. Tools such as R will allow them to do so.

No comments: