You are here
What's New in Revolution R Enterprise 7
The new Revolution R Enterprise 7 brings the latest performance-enhanced R engine to the largest data and most powerful compute environments. With this release:
- IT teams can run R analytics inside Hadoop clusters and Teradata databases, slashing model development and deployment cycles;
- Data Scientists can access a broader range of Big Data advanced analytics techniques, for more powerful predictive models;
- Business Analysts can use the intuitive drag-and-drop Alteryx interface, accessing the power of R without the need for programming.
Here’s what’s new in Revolution R Enterprise 7:
Upgraded R 3.0.2 Language Engine. The latest release of the open source R language engine brings improved performance and big-vector support to R, and compatibility with more than 6000 community-contributed R packages.
Write Once, Deploy Anywhere. With a single additional line of R code, re-deploy your Revolution R Enterprise ScaleR data analysis from a development server to run on a production grid or cluster, in Hadoop, or in-database. The following compute contexts are now supported:
- Hadoop: Revolution R Enterprise ScaleR now supports in-Hadoop processing for Cloudera CDH3 / CDH4 and HortonWorks HDP 1.3.
- Teradata Database [Available Q1 2014]. Teradata is the first enterprise database appliance to support in-database processing with Revolution R Enterprise ScaleR.
Drag-and-drop data analysis with Alteryx [Available Q1 2014]. Business analysts now have the power to easily build high powered analytics models with Alteryx, a desktop-to-cloud workbench. This integration brings the ability to create Big Data Big Analytics to users who are not R developers, but need access to the analytical capabilities of Revolution R Enterprise.
New Big Data Big Analytics Techniques. Revolution R Enterprise puts even more powerful Big Data Big Analytics techniques in analysts’ hands to generate and visualize the most impactful predictions and inferences:
- Ensemble Models for Decision Forests is a powerful tree-based machine learning technique similar to Random Forests®. (Random Forests is a trademark of Salford Systems.)
- Interactive Decision Tree Visualization. Decision tree visualization capabilities make it easier for data scientists and analytic consumers to understand relationships and correlations within the data.
- New Stepwise Regression Techniques. Stepwise Regression is now also available for logistic regression and Generalized Linear Models (GLMs), in addition to the previously-available linear regression. Data scientists now have more variable selection techniques available for a broader range of regression models and data types.
New Accelerators for Enterprise Deployment: Integration “accelerators” — examples built on the Revolution R Enterprise DeployR Web Services framework — provide a starting point for custom R integrations with Tableau and Excel. (These new accelerators are additions to existing examples for Jaspersoft and QlikView).
Improved ODBC Data Connectivity: Revolution R Enterprise ScaleR’s ODBC connection capability has been validated with the following data sources: HP Vertica, IBM DB2, IBM PureData System for Analytics, MySQL 5.1, Oracle Express, Pivotal Greenplum, PostgreSQL, SAP Sybase IQ, SQL Server 2008, SQLite 3 and Teradata Aster.
PMML Model Deployment: Predictive models created with Revolution R Enterprise ScaleR can now be exported to PMML for real-time model scoring.
Multi-Node Package Manager for Hadoop: new installation tools make it easier to deploy R code on Hadoop platforms.
New Server Platforms: Revolution R Enterprise Server is now supported on Windows Server 2012 and SUSE Linux Enterprise Server.