Hi, I have two visuals coming from different datasets with different granularity. The first one is in appointment level while the second one is more detailed in expense level. In the first dataset I have some ports and two columns named (col1, col2) boolean.
Now I want the second visual to show by default only the rows for the ports where the first visual has col1=‘No’ and col2=‘Yes’
So in the below example I want the second visual to show only St. Charles as in the first is No / Yes and not the other ports.
The simple answer is no because those visuals live on two different datasets. The way you could do this in Quick Sight would be with CustomSQL, but not in the analysis layer.
Hi @pantelis,
The only way to have fields from separate datasets interact with each other in Quick Sight, would be through a parameter that handles a ‘like’ field. So in this scenario, I don’t believe you would be able to achieve the proper filtering without a table join. You could setup an ifelse that checks if each field contains the proper ‘No’ and ‘Yes’ combination; then use that field to filter. But as that calculated field will be located within one dataset, it would not be able to filter the other dataset (visual 2).
Interactivity between two or more datasets is a common request so I’ll mark this as a feature request to promote visibility to the support team.
My team’s solution to this kind of problem was doing a SQL join to create a single dataset. How you do this highly depends on your existing data input infrastructure and the data itself.
In our case, we are working with CSV and Parquette files stored in S3, and accessed by Quicksight through an Athena datasource. Out in Athena, we create ‘views’ that join one or more tables of data into a useful single query, and then create a new Quicksight dataset from that view.
This can also be done with joins at the dataset level within Quicksight, but doing it as a view in Athena with a custom SQL query / join is much more powerful.
