How I fell in love with Tableau’s Relationships data model
An exploration of Tableau’s logical and physical layer data models
After reading Kirk Munroe’s blog post on shared dimensions I realised just how out of touch with Tableau’s logical data layer I was. So I went and read, watched and listened to everything I could about Relationships in Tableau and I can’t understand why anyone wouldn’t use it. Seriously, how can you resist it’s charms?!
Our first date didn’t go so well
I first started learning Tableau back in 2016 when making joins and unions, in the physical layer, was the only option. As much as any Tableau user likes to stay on top of the latest features in Tableau, occasionally things don’t quite make it into your work stream. Especially if you have already learnt a way to deal with x or workaround for y.
I remember seeing the demo of “Noodles” (now aka “Relationships”) at the Tableau conference and being a little sceptical as Tableau would be deciding the granularity of your data based on the data you put on the canvas. It seemed a little bit too much like magic and I no longer had control or visibility of exactly how the data was being structured. So I just stuck with the ol’ joins.
