Many to many without duplicating data?

Is there a way to handle many to many rlxn without duplicating data? I’ve seen suggested solution of a bridge table but I might be missing something basic here.
In the end on quicksight you can only have one big table so you will be joining your fact tables into the bridge table and duplicating the values.

How do you resolve that?

For example table even on a one to many
Customers and orders
If you have facts on customers table like age or amount spent on another service

These values will get duplicated when joined with orders table.

Is there a way to resolve this? One possible way is to just keep them seperate tables and have seperate charts and not mix the facts together but that sounds limiting.

@ame54 - As per my knowledge, currently there’s no way to handle many to many relation without duplicating the data at the dataset level. What we typically do, it create calculated fields using various LAC functions to get the correct/intended/de-duplicated results.

Since, Quick Sight doesn’t provide any way to model the data with defined relationships while keeping the tables separate, the bridge table solution will also not going to work at least at this point of time. Hope this helps!

2 Likes