Concept | Flow#

Watch the video

In Dataiku, the Flow is the visual representation of how data, recipes, and models work together to move data through an analytical pipeline.

Note

Dataiku positions the different items dynamically, in an optimized way. You cannot change the layout manually.

From the initial data to the final output, the Flow in Dataiku allows you to trace the dependencies among the different items and becomes a visual narrative of your data’s journey.

Screenshot of a Flow in Dataiku.

Improving the Flow readability#

Sometimes, the Flow can be quite complex, which can impact readability.

In such cases, for more clarity, you can use Flow zones, tags and filters.

Using Flow zones#

By default, a Flow is displayed in a single zone.

At any time, clicking the + Zone button at the top right of the Flow allows you to add zones to the Flow in order to organize the objects.

For more information, see the Tutorial | Flow zones article.

Screenshot of a Flow using Flow zones.

Tagging the Flow items#

When there are too many objects on the screen, you can add tags to the different items of the Flow then use these tags to select which parts of the Flow to view. Tags can be based on attributes such as creator, purpose, status and so on.

Example of a Flow with tags.

Filtering the Flow#

The View menu from the bottom left corner of your Flow lets you filter the Flow based on different elements such as Flow zones, tags, connections, recipe engines, the last modification date, etc.

For example, you can show or hide parts of the Flow by selecting or unselecting some tags to decrease the overall number of objects that appear on the screen.

Example of a Flow filtered with tags.

For more information, see the Concept | Flow views.

Building the Flow#

The Flow Actions dropdown at the bottom right of the Flow includes a Build all menu to build the entire Flow.

Yet, you can also right click on any item in the Flow and select the Build menu from there.

Keep in mind that the Flow in Dataiku has an awareness of the relationships and dependencies between datasets in your project. For example, if you make changes to a dataset or recipe, you may choose to dynamically rebuild dependent items upstream or downstream in your Flow to reflect the update.

For more information, see the Concept | Build modes article.

What’s next?#

In this lesson, you learned about the Flow in Dataiku. Continue getting to know the basics of Dataiku by learning about computation engines.

See also

For more information, see also: