Release Notes

Version 5.0.0 - December 8, 2017 (Preview)

The following release is a major feature update, which is available in preview for selected customers.

Features

Platform interface design update

  • We’ve updated the look and feel of our interface to improve the workflow and interactions across the Platform.

Zeppelin as an Interactive Session tool

  • Users can now launch Zeppelin sessions to explore their data via interactive data visualizations, pivot charts, and forms with built-in support for multiple languages such as Spark and SQL.

Built-in Cloudera Hadoop Support for Hive and Spark

  • For customers with Cloudera distributions of Hadoop, the DataScience.com Platform provides an easy form-driven method of configuring the cluster connection and installing all of the necessary dependencies.

Kerberos authentication via keytab upload

  • Users can upload their personal Kerberos credentials to connect securely to their external data sources.

Run and Publish reports

  • Prior to creating a report version, users can opt to run their notebooks top to bottom and publish the output, thereby streamlining the report workflow.

Version 4.2.2 - October 4, 2017

The following release is a minor feature update, ready for installation on the available release channel.

Features

Select files to sync

  • This feature is previewed in this release. From inside a Jupyter session, a user can select a subset of their modified files to sync to their project’s repository.

Resource Management for Users

  • Users will be informed that their compute resource size selection is not available to select due to current resource constraints on the cluster.

Version 4.1.1 - September 20, 2017

The following release is a minor feature update, ready for installation on the available release channel.

Features

Built-in MapR Hadoop support for Hive and Spark

  • For customers with MapR distributions of Hadoop, the DataScience.com Platform provides an easy form-driven method of configuring the cluster connection and installing all of the necessary dependencies.

Version 4.0.1 - September 6, 2017

The following release is a major feature update, ready for installation on the available release channel.

Features

Environment Management

  • Admins can now create and distribute customized, pre-installed collections of dependencies and packages as Environments to Users on the Platform.

Sync and Shutdown from Jupyter sessions

  • It is no longer necessary to switch tabs between your work and your Session Details page to sync or shut down your session.

File path autocomplete

  • Autocomplete your file path when running a script, publishing a report or app, or deploying an API

Enhanced Platform availability

  • Improved availability of the Platform application, Load Balancer, and Database

Single Sign On with SAML 2.0

  • Integrate with your SAML 2.0 provider for authentication

Version 3.9.1 - August 23, 2017

The following release is a minor feature update, available on the Stable release channel.

Features

Report Versioning

  • Users can now create multiple versions of a report under the same URL.
  • Users can also edit report version titles and descriptions.

User-supplied custom tagging for Amazon EC2 on-demand resources

  • End Users can now input custom tagging to Amazon’s EC2 metadata when provisioning on-demand resources.

Version 3.8.1 - August 9, 2017

The following release is a minor feature update, available on the Stable release channel.

Enhancements

  • Support for Gitlab version 7+ with API version 3
  • Support for Redhat Enterprise Linux version 7.3
  • Avatars with profile images
  • Minor bug fixes and security enhancements

Version 3.7.1 - July 26, 2017

The following release is a minor feature update, available on the Stable release channel.

Features

R Shiny dashboards deployable to the Outputs page

  • Users can now publish R Shiny applications to a dedicated Shiny server on the Platform, then share links to applications with project collaborators and business stakeholders.

Resource Management Dashboard

  • Admins can now manage all the server and Docker container resources running in the Platform: monitor RAM and CPU usage, identify unhealthy servers or analyses, and shut down unwanted processes.

Version 3.6.1 - July 13, 2017

The following release is a minor feature update, available on the Stable release channel.

Features

H2O.ai Dependency Collection

  • This dependency collection has H2O and its dependencies pre-installed to improve your AI capabilities

Enhancements

Enhanced Support for Internet Explorer 11

  • Improved experience of operating the Platform on the latest version of Internet Explorer

Version 3.5.1 - June 28, 2017

The following release is a minor feature update, available on the Stable release channel.

Features

Administrator-configured compute resources sizes

  • From the Admin Console, Administrators of the Platform can now configure the compute resource size options that are made available to Users of the Platform.

Various user experience and usability enhancements

  • Users can now name their sessions to distinguish them more easily from each other.
  • On session shutdown, users will be alerted to any un-synced changes in the session to prevent unintentional loss of work.
  • On a Run Details page for any past run, users will have the option to “rerun” the script at the latest commit, enabling users to quickly iterate while maintaining environment configurations.

Version 3.4.1 - June 15, 2017

The following release is a minor feature update, available on the Stable release channel.

Features

Curated Dependency Collections

  • Get started quickly on difficult data science problems with new dependency collections. New dependency collections now include the Standard set of packages as well as curated powerful packages for solving specific data science problems.

    • RStudio: Time Series
    • Jupyter: Deep Learning, Bayesian Analysis, and NLP

Multiple language kernels available in Jupyter sessions

  • Python 2.7, Python 3.5, and R 3.2 are available in each session.

Version 3.3.1 - June 7, 2017

The following release is a minor feature update, available on the Stable release channel.

Features

GitHub Enterprise and GitLab Enterprise integrations

  • In addition to GitHub.com, Bitbucket.org, and GitLab.com, the latest release also supports the enterprise versions of GitHub and GitLab for version control, allowing you to base your projects off of work in those repositories.

Global Environment Variables

  • Platform Admins can now set environment variables at the global level so that secrets needed across projects can be set once and managed in one place. Each key-value pair can have User- and Team-level permissions, ensuring control and security.

On-demand compute resources in AWS VPCs

  • For installations in customers’ Amazon VPCs, you can now control your cloud footprint and therefore your costs by provisioning single-use compute resources for sessions, runs, and deployed APIs. From the Platform, spin up an EC2 instance of your choosing, do your work, and shut it down when you no longer need that machine.

LDAP

  • Admins can connect their company’s active directory user management system to the DataScience.com Platform. LDAP-enabled environments use the specified active directory to authenticate users. Admins can manage users in their referred central location.

Version 3.2.1 - May 31, 2017

The following release is a minor feature update, available on the Stable release channel.

Features

Bitbucket.org and GitLab.com integrations

  • In addition to Github, the latest release also supports Bitbucket.org and GitLab.com for version control, allowing you to base your projects off of work in those repositories.

RStudio

  • In addition to Jupyter, users can launch RStudio interactive sessions where they can import pre-installed packages, use environment variables, and sync directly to their repository.

Publish RMarkdown HTML docs

  • After creating analyses in RStudio, publish your findings for your teammates and colleagues as sharable, reproducible reports.

Version 3.1.1 - May 4, 2017

The following release is a major feature update. It is now available as version 3.1.1.

Features

Projects

  • Ensure visibility of your team’s work by organizing code, data, and outputs into projects. Centralize knowledge, invite collaborators, and do work that is focused on solving the problems that impact your organization.

GitHub integration

  • Version your shared code and watch your projects progress. Track decisions, milestones, and project lineage over time, and never lose sight of your work again.

Secret management

  • Avoid committing secrets to code by storing them as secure environment variables.

Launch Jupyter Interactive sessions

  • Quickly spin up a Jupyter interactive session with support for Python 2, Python 3, and R.

Publish Reports

  • Turn Jupyter notebooks, markdown documents, and other files from your project into reproducible reports that can be shared across your organization.

Deploy APIs

  • Deploy Python and R code behind a REST API to make it instantly available for integration with real-time applications or dashboards.

Run scripts

  • Run Python and R scripts from a web UI and share outputs across your team.

Schedule Runs

  • Run code on a schedule to automate data science processes.