The overwhelming majority of the analysis I do in undertaken in a statistical program language called R, an open-source application that is maintained, curated, nurtured, loved and continuously developed by a vast network of volunteers.

For simpler tasks and for presentation purposes I use Microsoft Excel which, especially with the enhancements that came with the 2010 version, is a reliable workhorse for garden-variety analyses.

When I believe that there's a signal to be found in some dataset but I'm not sure which variables are the semaphores and which the ciphers, and when I'm uncertain about the exact nature of the relationship between predictors and target, I call on the services of Eureqa Formulize to spot me a few clues. It's also open source and, while it's been impressive since its first release, has become slicker with every version.

From time-to-time I've also experimented with Phil Brierley's Tiberius Data Mining tool. 

Last updated on January 4, 2013 by TonyC