Microsoft Updates R Open Programming Language, an AI and ML Favorite
Microsoft has issued a new release of its enhanced edition of the R programming language, one of the primary languages used for artificial intelligence and machine learning development, along with Python.
Although R is also used for many other tasks, it has secured a place atop the favorites lists of many AI programmers. As Wikipedia states: "R is widely used in new-style artificial intelligence, involving statistical computations, numerical analysis, the use of Bayesian inference, neural networks and in general machine learning."
Earlier this month, Microsoft updated its R Open offering, with improved multi-processor support highlighting new functionality in the new release.
Microsoft R Open is the company's enhanced, open source distribution of R, commonly used for statistical analysis and data science.
The company this month updated its offering to version 3.5.1, which is completely compatible with R 3.5.1, which shipped last month.
Highlighting new functionality coming with the update is new support for multi-threaded computations via new math libraries.
"These libraries make it possible for so many common R operations to use all of the processing power available," Microsoft said. "Matrix operations, most notably, can compute in parallel with all available processing power to significantly reduce computation times."
The company said they automatically use all available cores and processors on a machine. "There's nothing special any package needs to do to benefit from these libraries either," Microsoft said. "In fact, any package used for vector/matrix operations will experience enhanced performance automatically when these libraries are installed."
Another key enhancement highlighted by the R Open team is a high-performance default, fixed CRAN (Comprehensive R Archive Network) repository. This post explains how it provides a consistent and static set of packages to R Open developers and users. It's based on a fixed CRAN repository snapshot dated Aug. 1, 2018, and won't change until the next release.
Also new is a checkpoint package. "The checkpoint package is also installed from CRAN during the installation of Microsoft R Open," the company said. This package, available on CRAN, is designed to make it easy to write reproducible R code by allowing you to go backward (or forward) in time to retrieve the exact versions of the packages you need. All you need are two lines of code to access package versions from a different date." More information on that feature can be found here.
Many new R packages are also available in that Aug. 1 snapshot, contributed by the community surrounding the open source offering. There are dozens of packages available for applications, data munging, data sources, graphics, image and natural language analysis, interfaces, machine learning, programming tools and more.
David Ramel is the editor of Visual Studio Magazine.