Creating a scatterpolt that updates based on a "dynamically" populated table

[Anonymous]

Hi everyone,

I am trying to make a scatterplot that updates dynmically based on a filter selection and uses data points pulled from two columns. Here's a small dataset similar to what I'm working with.

Client	Industry	Index A	Index B
Bubble Master	Soap	3	2
Sud King	Soap	4	5
Soapy Mess	Soap	2	3
Clean Wash	Soap	3	2
Meaty Kibble	Dog Food	1	1
Dog's Delight	Dog Food	1	1
Tasty Treat	Dog Food	1	1
Woof Dinner	Dog Food	1	1

I would like to create a scatterplot that maps Index A on the X-axis and Index B on the Y-axis, has a data point for a single client (based on a filter selection), a data point for the Industry associated with the Client (mapped on the averages of Index A and Index B), and an "Overall" data point for the average Index A and Index B for the entire dataset. Here's a very barebones example of the scatterplot, with the Sud King selected as the Client.

If a different Client were selected, I'd like the data points to change (e.g., if Tasty Treat were selected, then both the Client and the Industry data points would move). I've tried making this work by contructing a table that pulls from the dataset, but Power BI cannot make dynamically updating tables. 

Does anyone have a suggestion re. what to do? Thanks for any help.

Cheers,

T

Update: Please note, for the data points in the example scatterplot, the size of the bubbles/dots is not relevant. The bubbles are simply illustrative and could be replaced with equal-sized dots. Also, the coordinates for the Industry and Overall data points are calculated using the mean values of Index A and Index B. (Thanks for the suggested updates, lbendlin.)

[lbendlin]

Great question, similar to "Top 5 and Others"  Showing the top 5 products and Other row - SQLBI

However, there are a lot of caveats here.  Firstly, your data is missing the weight column (the bubble size).  And then - what is your strategy for calculating "Soap"  - as an average? as a Least Squares interpolation? Some other grouping algorithm?