I'm not exactly clear on the visualization you are expecting. You can segregate the data using Slices and then use each Slice in a separate chart.
If you are asking if there is a way to aggregate values on the fly and then use the aggregated values in the chart, the answer is no - not directly. The values in the scatterplot would need to be stored in some column. The preferred way is to use a separate table the summarizes the data as needed then used in your charts. This extends very well as the app data grows and keeps any performance-impacting processing minimized.
Some have used Virtual Columns. The idea is that each row in the group gets the same calculated aggregated values but then you select a representative row (maybe with a Slice - which can ALSO select the specific columns needed) that is used in the chart. The issue with this approach is that the Virtual Columns are re-calculated on EVERY row with EVERY data change or Sync. If you have a large table or have many such Virtual Columns, Sync times will be slowed and in several situations app performance itself can become slow. Additionally, this approach doesn't extend well. If new groups are needed you will be required to make manual app changes to include them.
I hope this helps!
LinkedIn
Twitter
