What's new in Spotfire® Data Science - Team Studio

TIBCO® Data Science - Team Studio 7.1

TIBCO® Data Science - Team Studio version 7.1 has been released with the following updates:

Updated workflow engine

This latest release uses Spark 3.3 for better optimization and faster execution of workflows as single Spark applications. PySpark support is also updated to Spark 3.3 for Python notebooks.

New Spark operators

TIBCO Data Science - Team Studio 7.1 introduces 2 new operators for the updated workflow engine (Isolation Forest and Import Excel). 

Performance Enhancements

This release includes a number of optimizations for remote data sources, workflow executions and notebooks.

TDV Integration Enhancements

This latest release has an easier TDV - TIBCO Data Science - Team Studio integration setup process.

Numerous Bug Fixes

We are happy to share that we have included fixes for a number of customer-reported issues in this release.

Learn More

• TIBCO® Data Science Team Studio version 7.1 documentation is available here and includes release notes. 
• TIBCO Data Science - Team Studio Enablement Hub

TIBCO® Data Science - Team Studio 7.0

TIBCO® Data Science - Team Studio version 7.0 has been released with the following updates:

Containerized for easy and more flexible deployment

TIBCO Data Science Team Studio 7.0 is re-architectured as a set of docker containers. This means that it can be deployed into any environment supporting docker. This includes on-premise and cloud environments.

Fig 1: Containerized Architecture

New and improved workflow engine

This latest release comes with a brand new workflow engine which allows for faster execution of workflows as single Spark applications. It uses Spark 3.2 / MLLib  for better optimization and faster execution. 

Following the introduction of TIBCO Data Virtualization (TDV) with version 6.6, TDV is now fully integrated for Data Management and Data Access to non-spark data. Computations can be distributed in-cluster and in-database for handling Big Data use cases.

New operators, extending Data Science capabilities

TIBCO Data Science Team Studio 7.0 introduces 44 new operators that exploit the new workflow engine and TDV integration. New operators such as Wide data selection, dynamic variable selection, and self-organizing maps are designed specifically for the High Tech Manufacturing Industry.

Improved Python execution performance

With TIBCO Data Science Team Studio 7.0 we have taken several steps to improve speed and scale of performance of Python execution.

Integration with TIBCO Spotfire and TIBCO ModelOps

TIBCO Data Science Team Studio 7.0 is fully integrated with TIBCO ModelOps for the execution of Spark Models created in Team Studio directly in your production environment, with all the benefits of TIBCO ModelOps. Model operations include a Spark runner for scalable inference and traceable management and governance of model artifacts.

Learn More

• TIBCO® Data Science Team Studio version 7.0 documentation is available here and includes release notes. 
• TIBCO Data Science - Team Studio Enablement Hub

TIBCO® Data Science - Team Studio 6.6

TIBCO® Data Science - Team Studio version 6.6 has been released with the following new features:

Support for TIBCO® Data Virtualization

TIBCO® Data Science - Team Studio version 6.6 can join forces with TIBCO® Data Virtualization (TIBCO® DV) in order to connect to a diverse set of data sources, including cloud data sources such as AWS S3, while maintaining the machine learning capabilities of Spark 2.4. To enable this, we provided 19 new operators (Mods) designed to work with TIBCO DV data sources. This allows users to build a complete data science workflow whilst still being able to use standard database operators provided by TIBCO Data Science - Team Studio. The SQL queries are pushed down to TIBCO DV, and the machine learning executes in Spark. 

The new operators are provided on the TIBCO Community Exchange, and each operator is individually documented. The result is a set of scalable operators capable of processing large volumes of data.

Further Features

• User interface improvements to dialog boxes
• Enhanced usability of Parquet files
• Ability to select multiple Hadoop files in the same operation
• Support for running temporary functions in Big Query using SQL Execute
• Re-engineered PCA operator
• Additional operators (Wide Data Variable Selector, Time Series SAX Encoder)

See the latest example of TIBCO Data Science - Team Studio 6.6 in this TIBCO Analytics Meetup video:

TIBCO® Data Science - Team Studio version 6.6 documentation is available here and includes release notes. 

Note that a TDV data source must be connected to the individual workflow file to be able to use the TDV operators. 

TIBCO® Data Science - Team Studio 6.5

Starting with release 6.5, the product's name is changed from TIBCO Spotfire Data Science to TIBCO Data Science - Team Studio. 

Product documentation can be found on the TIBCO Documentation website including Release Notes.  

TIBCO® Data Science - Team Studio 6.4

TIBCO Data Science - Team Studio has a new release, version 6.4. In this release, you will find several new operators, fundamental performance improvements in data preparation steps, and various user experience enhancements. All new features and improvements are summarized and documented in detail in the Release Notes.

We're particularly excited about Node Fusion for Hadoop-based workflows. Node Fusion means that multiple operators can be run as a single Spark job which makes computation significantly faster - sometimes by order of magnitude. The screenshot below illustrates how all 8 purple-colored operators in the workflow are run together in this single Spark job: 

The new release also extends the integration possibilities of TIBCO Analytics products. In particular, analytical models developed in TIBCO Spotfire Data Science can be now automatically sent to TIBCO Streaming where these models can be utilized for real-time scoring.

• Changing from Tables to Views: You can now make bulk changes to whether your database workflows are outputting Tables or Views. 
• New operators, including:
  • Chi-Square, Independence Test
  • N-gram Dictionary Loader
  • One-Hot Encoding
  • Reorder Columns
  • Sessionization
  • Chi-Square, Goodness of Fit
• Many other improvements to usability and performance 

For additional information on Spotfire® Data Science - Team Studio, search previously answered questions and post your questions in Forums. Use the Spotfire® Data Science - Team Studio Enablement Hub to learn more.