Prepare and Transform Data

Recipes are the fundamental building blocks of a Dataiku project. Every time you perform a transformation (such as joining tables, aggregating rows, or cleaning values), you create a recipe in the Flow.

At its core, a recipe defines the logic used to produce one or more output datasets from one or more input datasets. Recipes can be:

Recipe type	Use case
Visual	Perform common data operations (like join, group, or prepare) using a point-and-click interface. No coding required.
Code	Implement advanced logic using Python, R, or SQL when standard visual steps don’t meet specific requirements.
Plugin	Access specialized components or wrap custom code in a reusable visual interface for other users.

Note

This section focuses on visual data preparation. You can learn more about code recipes and plugin recipes in their sections of the Knowledge Base.

Topics

• Discover Visual Recipes
• Process Unstructured Data
• Prepare Data
• Join Data
• Aggregate Data
• Enrich Data
• Reorder Data
• Transform Data
• Reduce Data
• Stack/Split Data
• Build Data
• Migrate from Other Tools