Design Datasets with Dataflows and the Dataset Builder

Use a dataflow to create one or more datasets based on source data from existing datasets or synced data. A dataflow is a set of instructions that specifies what input data to include, how to transform that data, and which datasets to load the transformed data into.

To build a simple dataset from related Salesforce data, use Dataset Builder. To build more complex datasets that can include external and transformed data, use Dataflow Editor. Both tools add the logic for building the datasets to a dataflow.

After you design a dataflow with either tool, run the dataflow to create the datasets. You can schedule a dataflow to refresh the data in the datasets on a regular interval. You can create multiple dataflows if you want to run them on different intervals or break up the data integration logic used to build your datasets. Just keep in mind that there’s a limit on the number of dataflows you can create in your org.

• Why Should I Use Recipes Instead of Dataflows?
  Recipes and dataflows both prepare data. Compared to dataflows, recipes are newer and are recommended for functionality and ease of use. Recipes allow you to preview the data as you transform it, while dataflows only show your node schema. For example, recipes have more join types and transformations with built-in machine learning - such as Predict Missing Values and Detect Sentiment - that aren’t available in dataflows. Recipes can also aggregate data to a higher level.
• Convert a Dataflow to a Recipe
  To use joins, aggregates, and sentiment detection in Data Prep, convert your dataflows to recipes. However, if your dataflow achieves your desired outcome, there’s no need to convert it, and you can continue using it. But consider using recipes for new Data Prep jobs because they offer more features.
• Design a Simple Dataset with Dataset Builder
  Use Dataset Builder to create a single dataset based on data from one or more related Salesforce objects. Dataset Builder adds the instructions for building the dataset to the specified dataflow. The dataset is created the next time the dataflow runs, and refreshes each time the dataflow runs thereafter. To edit the dataflow, use the Dataflow Editor.
• Design Complex Datasets with Dataflow Editor
  Use Dataflow Editor, a point-and-click interface, to build your dataflow logic from scratch. Add transformations to determine what source data to use, how to transform that data, and which datasets to load the results into.