High performance statistical computing with parallel R: applications to biology and climate modelling

Abstract

to massive data sets. To address this challenge, our tactical approach is based on parallelizing a statistical package, called R Similar to IDL, R is a non-parallel language and is not designed to handle massive data-sets. Our choice of R, however, is driven by its open source availability, extensibility, and very broad usability by a large scientific community. Our long-term strategy is to provide a middleware between a specific high-level data analytics language such as R, IDL, or MATLAB and a generic, parallel, optimized, portable computational analytics engine.