Previous Releases

Analytics Framework Service

Q2 2020 Release

Obsolete Features

The following features are no longer available.

Python versions older than V3.X are no longer supported for developing analytics for the Predix Analytics Framework service. You must now develop new analytics using Python 3.X. However, you can successfully run the analytics developed using Python 2.X.

Tip: GE Digital recommends you migrate the analytics developed using Python 2.X to Python 3.X. For more information on how to migrate the analytics, refer to the Python documentation.

Q1 2018 Release

Enhancements

The following new enhancements have been added.

Support for analytics using Predix Time Series data having separate timestamps

Hierarchical analytic template type supports separate timestamps for each indexed block of generated data for variable input port. For more information, see:

New field definition (timestampShared) for FieldPort block in afs-reference.html#reference_eqd_fvr_rv.
New sample (Analytic Requiring Predix Time Series Data Using Variable Input Ports) added to afs-reference.html#reference_ed188d2e-9dbe-46f1-8ec6-93aecea8f8b4.

Support for analytics using Predix Time Series data with array of tags

You can run analytics using Predix Time Series data that is an array of tags. For more information, see:

New sample (Analytic Requiring Predix Time Series Data Using an Array of Tags) added to afs-reference.html#reference_ed188d2e-9dbe-46f1-8ec6-93aecea8f8b4.
afs-orchestration-execution.html#task_9d270d7f-4d4a-4e07-9239-f121d17f1c8d.

Q4 2017 Release

Enhancements

The following new enhancements have been added.

Support for Parametrized CONSTANT as a Custom Attribute: A constant input value can be parametrized and passed in at runtime through the customAttributesfield of OrchestrationExecutionRequest in the port-to-field map. For more information, see afs-orchestration-execution.html#task_0144efe9-9c83-459f-8d56-2594b90f83f2 and afs-reference.html#reference_vqh_ms2_sv.

Q3 2017 Release

Enhancements

The following new enhancements have been added.

A migration API is now available to move catalog and orchestration data from your original analytics service instances to your new Analytics Framework instance. For more information, see Task Roadmap: Migrating to Analytics Framework Service.
If you currently have an Analytics UI UAA Client using implicit or authorization code grant_types, you must configure your app with the correct redirect_uri. Update your existing analytics UI UAA client redirect_uri value to <your_analytics_ui_url>/callback. Failure to do this will block you from signing in to the Analytics UI. For more information, see:
- How-To: Add redirect_uri to your UAA client article on Predix Forum.
- afs-get-started.html#task_9144ad8f-1c20-40bf-8dbd-43ff69594f43.
- afs-get-started.html#task_ny3_mr5_vv.
Orchestration Performance Improvements
Improved Analytic Log Retrieval in Analytics Catalog

Q1 2017 Release

New Features

The following new features have been added.

Analytics Framework Service: The Analytics Runtime service is deprecated and being replaced by the Analytics Framework service. For more information, see Analytics Framework Service FAQ.; The Analytics Framework service consolidates into one service all the features and functionality previously offered by the three analytics services (Analytics Catalog, Analytics Runtime, Analytics UI) combined. The Analytics Framework service simplifies developing advanced business analyses and deploying them to business operations. For more information, see About Migrating to Analytics Framework.
Customize Query Criteria: Support for custom query attributes when using Predix Time Series as a data source has been added. For more information, see afs-reference.html#reference_47e1794f-1a98-4def-9b7e-3df56137cad7.
Dynamic Tag Support: Support for dynamic tags in port-to-field maps has been added. When running an orchestration request using Predix Time Series tags, you can pass dynamic tags using customAttributes as a key-value pair. Once orchestration execution is complete, the dynamically created time series tags are sent to the monitoring service from which you can query to read orchestration execution details. For more information, see afs-reference.html#task_476c42a9-99df-4703-b172-d7e5872b0415.
Descriptive Analytic Template: Support for descriptive analytic templates have been added. The descriptive template format simplifies the template structure as the data format is flat but has self-describing attributes such as data type and data size. The descriptive analytic template reduces the burden on the template developer by reducing the need to convert strings to specific data types required by the analytic. For more information, see afs-analytic-template.html#concept_09a2dec6-5f3e-4e05-90f9-a383fe2f6ad1.

Enhancements

The following new enhancements have been added.

Passing Custom Attributes During Orchestration: Support for passing custom attributes during an orchestration has been added. For more information, see afs-orchestration-execution.html#task_0144efe9-9c83-459f-8d56-2594b90f83f2.