Circle packing with multiple attractor points

In this definition each circle is treated as a single person. Each person belongs to a specific “center” which is represented by colors and also has a specific location. The idea is to use kangaroo to packing these people around their locations which is given by different points.

In given times the total population can reach highest than 3.000 people distributed on 23 differents locations (points) where each one is colored based in your respective center. It’s a lot of interactions and the only way I found to solve that is exploding the data tree in 23 branches containing the values to the mass addittion.
The only way to illustrate what I would like to do was making use by one solver for each location. Obviously, it’s a awful solution. But, the most important thing is the circles needs to packing around their respective points (location).

I’d like to get some thoughts about how to optimize this definition.

Thanks ;D

Cheers!

circlepacking.gh (35.2 KB)
results.rar (17.2 KB)