Feature sets
A feature set is a collection of features stored and versioned by Feature Store. These pages follow its lifecycle, from registration through versioning to maintenance. To load data into one, see Ingest data; to read it back, see Retrieve data.
Register a feature set
Register a new feature set from a schema, including derived feature sets, time travel columns, and masking of sensitive features.
Find and inspect feature sets
List feature sets in a workspace or across workspaces, obtain a specific version, read its properties, and preview its data.
Work with features
Set a feature's type and read its computed statistics.
Update feature set metadata
Update tags, description, deprecation state, time to live, and other fields on a feature set and its features.
Create a new version
A feature set is a collection of features. Users can create a new version of an existing feature set for various reasons.
Evolve the schema
Read a feature set's schema, check it against a new data source, and patch metadata across into a new schema.
Derived feature sets
Determine whether a feature set is derived, and trace its lineage to parent and child feature sets.
Review feature sets
Feature set review process involves the reviewer's acceptance. Depending on the system configuration, all feature sets or only sensitive ones may be subject to review.
Artifacts
List, retrieve, save, and delete the file and link artifacts attached to a feature set.
Maintain a feature set
Refresh a feature set, mark target variables, open it in the Web UI, and optimize its storage.
Delete a feature set
Delete a whole feature set, or a single major version of one.
- Submit and view feedback for this page
- Send feedback about H2O Feature Store to cloud-feedback@h2o.ai