DataRobot supports multiple methods of feature engineering— automatic and manual feature transformations for single datasets, as well as Feature Discovery for multiple datasets. Learn guidelines and best practices for using DataRobot's feature engineering tools.
|Transform primary datasets||Automatic and manual transformations, as well as interaction-based feature creation within a single dataset.|
|Feature Discovery with multiple datasets||Multi-dataset, interaction-based feature creation.|
The topics below include additional information on transforming data:
|Prepare data in AI Catalog with Spark SQL||Enrich, transform, shape, and blend together datasets using Spark SQL queries within the AI Catalog.|
|Shape data||Use data shaping tools to deduplicate, group, transpose, and pivot data.|
|Attach datasets||Perform lookup, join, and append operations on your dataset.|