Product area

Features

Feature engineering and the Feature Store in Unity Catalog: feature tables, lookups, point-in-time joins.

Concepts

4 in this area
  1. 1Feature engineering and the feature storeFeature Engineering in Unity Catalog stores reusable model inputs — feature tables — as ordinary Delta tables that carry a primary key, plus a Python client, FeatureEngineeringClient, that knows how to join those tables correctly when it builds a training set…
  2. 2Training sets and point-in-time joinsA training set is the object fe.createtrainingset() returns: a declaration of which feature tables to join to a labelled DataFrame, and on which keys. trainingset.loaddf() materialises it, and fe.logmodel(..., trainingset=trainingset) stores the declaration…
  3. 3Online Feature StoreAn Online Feature Store is the low-latency half of the feature store described in feature engineering. An offline feature table is an ordinary Delta table in Unity Catalog: columnar, cheap to scan, and the wrong shape for fetching one row by key inside a…
  4. 4Feature ViewsA Feature View is a feature described rather than built. Instead of writing a job that aggregates a source table and writes the result to a feature table, you declare four things: a source, an entity to group by, a timeseries column to order by, and a…

News

No news for this area.