Tags & Object Descriptions

Find resources for the best utilization of tags and object descriptions for greater collaboration in Dataiku.

Tutorials

• Tutorial | Tags

Tip | Suggestions for using tags

Using tags extensively in Flow helps identifying at a glance the role of each parts of the Flow. You can also tag elements with the name of the person that is responsible for it! Tag color can be changed (use for instance red tags for important or urgent elements):

Suggestions for good tags:

• Thematic tags: tag branches dedicated to specific tasks (e.g. “insights”, “preprocessing”), tag inputs as “sources”…

• Tag the parts of your Flow that are scheduled to run automatically, so you know that changing them will affect a production workflow!

• Tag with the progress status: work in progress, done, in production…

• Tag with another person’s name to draw their attention to that part of the Flow.

Tip | Commenting to document Dataiku objects

Adopting a good dataset naming scheme avoids a lot of long descriptions and many comments. However comments are still very useful when collaborating with others (note that yourself in six months counts as someone else).

There are many places where a few words can be very useful:

• A description on the project homepage. You can add links to datasets, recipes, or any element of the project.

• A description in the “summary” tab of a dataset. Note this appears on the flow: click the dataset then “details” in the right column

• When publishing insights, you can add the link of the corresponding dataset in the description text.

• Edit column details to add a short comment.

• Comments in the code of your custom recipes.

  • There are many online resources about writing good comments. Just a few words:

    • Explain what you intend to do, what it will be used for. (How it’s done and what a recipe does is less important.)

    • Imagine that you’re taking the project over, what would you want to read here?

    • Time for some fun