Revolution Analytics releases version 5.0
Revolution Analytics has released version 5.0 of its commercial R-based enterprise distribution. The new version of Revolution R introduces key features for statistical analysis of "Big Data" by integrating the RevoScaleR "Big Data functions" for efficient handling of terabyte-sized data sets.
The enhanced RevoDeployR server extends the GPL-licensed open-source version of R with a RESTful web services framework which can be used to integrate the statistical functions into other applications. Revolution Analytics also sets out to significantly improve upon the performance of R with Revolution R Enterprise 5.0.
The new version adds the capability to distribute statistical analyses from a desktop and run it across the nodes of a cluster using Windows HPC server. This mechanism can also distribute R functions across nodes. New data management capabilities improve flexibility with added import and manipulation/cleaning tools.
The company is also introducing integration with Apache Hadoop through a partnership with Cloudera. The integration allows MapReduce programming in R and the ability to access HDFS and HBASE data. Revolution R 5.0 includes a fully-patched version of R 2.13.2, which features a new byte-compiler to improve performance of user-written functions and packages.
Revolution R version 5.0 is a commercial enterprise package but a single user subscription available free of charge to academics. Further details of the enhancements of the enterprise package are available in the blog posting announcing its availability. The GPL licensed version of the R language is available to download from the R Project's mirror sites.