Release Date: May 5, 2016
Updated: May 12, 2016
We are glad to announce another monthly ecosystem update release.
ECO-1604 release includes GA version of Apache Spark 1.6.1 for MapR (emerging from a 2-month Developer Preview) and updates to existing ecosystem packages.
*Update 5/12/16: Today, we have re-released Hive 1.0.0 with fixes as part of ECO-1604 release.
Be the first to get notified about new releases
ECO-1604 Release Features
See the Spark 1.6 Developer Preview announcement for all of the added features: Essential Tips for Making the Most of the Spark 1.6 Developer Preview | MapR
Automatic memory management: execution memory and storage memory are adapted dynamically based on workload characteristics.
Dataset API: Datasets can be constructed from JVM objects and then manipulated using functional transformations (map, flatMap, filter, etc) in Scala and Java.
Machine Learning Pipeline Persistence: adds new features to machine learning that takes persistence beyond models to persisting the entire pipeline, including transformers and estimators.
New Machine Learning algorithms: online hypothesis testing (A/B testing in Spark Streaming), bisecting K-Means clustering, and normal equation for least squares that provides R-like model summary statistics.
Ecosystem Package locations:
- Cross-Site Scripting (XSS) Vulnerability for Hue
- Release Notes
- Installation Instruction
Converge Community Resources
Have a Question?
Ask in Answers or comment below.