Relationships in Tableau have grown older – it’s been almost two years now when they started seeing the light of day. And still, even after all that time, there are many questions around this data model, be it within the Tableau Community, be it in Tableau trainings or be it with clients who have been really comfortable with joins and blends in the old data model.
And first off: Thanks so much to Sean Spencer, Chris Hastie, Kent Sloan, James Austin, Helge Thomson, Raphael Teufel and Madeline Cook for helping out with feedbacks, critiques, clever questions and insights that I wouldn’t have found on my own. Also, thanks for some of the first responses that took my ask for relationship advice a bit too literal. 🙂
When Tableau dropped the news about the new data model during the “Devs on Stage” session during the Tableau Conference 2019, it was an atmosphere of pure excitement. The kind of excitement we have when we hear fascinating news, wonder in amazement at what’s coming and have no clue yet how it will work. No more joining tables? Drastically reducing the need for LOD functions? No more blends? Performance jumps?
It sounded too good to be true. But hey, Tableau delivered (to be fair, it took a few versions until all the gear wheels were interlocked). So, for people, who didn’t bother using the Noodle, or who have used it but never understood it, here is a crash course for you.
This mini-series will cover relationships in Tableau: what they are, how they work, what the different layers are, ground rules, basics, tips and tricks, and also a deeper dive into the performance options. As there is just too much to tell within a single article, we split the content into a bunch of articles. Here’s what we were up to:
- Basics: Logical Layer & Physical Layer
- Basics: The Ground Rules
- Tips & Tricks
- LODs
- Extracts
- Performance: A Few Introductions
- Performance: Speed and Cardinality
- Performance: Joins and Referential Integrity
- Performance: Breaking the Cardinality
My hope is that this series will help you utilize relationships in Tableau to the fullest extent. We may also add new topics as posts as they emerge as the world of relationships is big and always growing.
