I had recently created a tableau report which had data from multiple tables and used its “relationship” feature to be able to link the data at time line level and create calculated fields. This functionality is missing in QuickSight, and having to create multiple cross joins with the given list of data sets is not showing up the right numbers in the visual. Is there any work around for this? Or is it suggested to have multiple datasets created on QS where each dataset will focus on 1 level of calculation? Thanks.
Hi Gauri, are you able to explain your use case with an example?
At the moment QuickSight does not provide the “relationship” feature (which Tableau has added in the last 2 years, if I remember correctly). At the moment you would need to define multiple datasets with different joins, depending on the data used in different visuals.
We have discussed providing similar functionality in QuickSight.
To try revive this thread, we are finding huge limitations in the complexity of analysis Quicksight can manage vs tools like PowerBi/Tableau with the sole use of joins between datasets (vs relationships in those other BI solutions).
To map together multiple data types, and analyse across multiple variables, we end up with exponential growth of the final data table with hundreds/thousands of duplications.
While the use of Aware Level calculations (covered in a recent training video provided by AWS quicksight team) fills some of the functionality gap, it needs constant checking & pretty advanced calculations to ensure the right results are being reported.
I’m sure all the AI features being added for insights is great, but if we have to constantly battle this constraint with complex work-arounds or simplify the data/analysis needs, then we’ll never end up leveraging them (and will ultimately need to move back to those other solutions).
Would someone mind letting me know if adding this functionality did make it to the planned roadmap for Quicksight, and if so what is the expected timeline?
Very happy to provide examples (in private) if useful to clarify further.
Thanks & hopefully this is useful feedback.
I am also having this issue with just a simple join I am getting lots of duplications…