Table structure for making a dynamic and efficient scatter chart visual using a very large dataset

Hey Anonymous ,

your question is quite generic, and you might consider to provide a pbix that contains sample data and reflects your data model, upload the pbix and xlsx (if you used xlsx to create the sample data) to onedrive and dropbox and share the link.

Nevertheless, create slicers that allow for selections what should be shown on the x-axis and y-axis. The content for this slicers is derived from tables that are in no way related to your existing tables.

Assuming that one of theses tables is called "xaxis" and has a column called KPI, you can create a measure like so:

KPI for xaxis =
var selectedkpiforxaxis = SELECTEDVALUE('xaxis'[KPI] , <[name of th default measure]>)
return
SWITCH
    selectedkpiforxaxis
    ,"name of 1 kpi" , <[name of the corresponding measure]>
    ,"name of 2 kpi" , <[name of the corresponding measure]>
    , ...
    ,"name of n  kpi" , <[name of the corresponding measure]>
)

Create a 2nd measure for the content of the yaxis.

Use both of these measure on the scatter plot.

Hopefully this provides you with some new ideas.

Regards,

Tom