Microsoft R Server 9.0, Microsoft's R distribution with added big-data, in-database, and integration capabilities, was released today and is now available for download to MSDN subscribers. This latest release is built on Microsoft R Open 3.3.2, and adds new machine-learning capabilities, new ways to integrate R into applications, and additional big-data support for Spark 2.0.
This release includes a brand new R package for machine learning: MicrosoftML. This package provides state-of-the-art, fast and scalable machine learning algorithms for common data science tasks including featurization, classification and regression. Some of the functions provided include:
- Fast linear and logistic model functions based on the Stochastic Dual Coordinate Ascent method;
- Fast Forests, a random forest and quantile regression forest implementation based on FastRank, an efficient implementation of the MART gradient boosting algorithm;
- A neural network algorithm with support for custom, multilayer network topologies and GPU acceleration;
- One-class anomaly detection based on support vector machines.
You can learn more about MicrosoftML at this live webinar on Wednesday, December 14.
The RevoScaleR package, which provides big-data support for Microsoft R Server, has been updated to support Spark 2.0 (This is in addition to existing support for in-database computations with SQL Server and Teradata.) New functions allow you to connect to a Hive, Parquet or Spark DataFrame data source, and execute data manipulation and predictive modeling tasks directly on the data as it resides within Spark.
Microsoft R Server also sports new capabilities to integrate R functions into other applications. (This is the results of enhancements to the former DeployR project, which is now part of Microsoft R Server.) Microsoft R Server can now host R functions exposed as web services: data scientists can publish R functions to the R Server, which can then be accessed from any application. Application developers can integrate those R functions into their applications using easy-to-consume Swagger-based APIs from any programming language.
Also released today is the free Microsoft R Client 3.3.2. This is a desktop edition of Microsoft R Server 9.0, designed for local data science development and remote execution on local servers and in the cloud. It includes the same MicrosoftML package as Microsoft R Server. The RevoScaleR package in Microsoft R Client is limited in data size when computing locally, but can shift computations to a remote Microsoft R Server when working with larger data sets. It also includes the new mrsdeploy package, so you can remote-execute any R code on the remote R Server, or even publish R functions as a Web service.
For a complete list of changes, see What's New in R Server 9.0.1. And for more on the release of Microsoft R Server 9.0, check out the blog post linked below.
Cortana Intelligence and Machine Learning Blog: Introducing Microsoft R Server 9.0