I have someone who has given me a thumbnail of a map with some concentric ellipsoids placed in estimated high density locations for our business. One location has three, and two other location have two and one respectively.
The assumption is that within the first ellipsoid is about 75% of our business, and within the second and third combined make up about 20% percent of the remaining.
They've asked me to create this, with numbers to back it up. The data I have right now is Zip Code Tabulation Areas for our geographical region, with registration information in a wide range of categories.
I've recently been utilizing GIS and delving into the world of spatial analysis, so forgive me if I'm naive. My thought was to find a technique that would allow me to fit an ellipsoid polygon around the polygons or their centroids for the density of ZCTA's that represent the 75% respectively, and create another one encompassing it to capture the remaining 20%. Those would be the areas we would focus on most heavily.
The question is... how would I even approach this in an application like QGIS? Furthermore, how would I break out the problem so that the two separate regions make up some of the calculations of that 75% and 20%.
Is this just something I have to approximate? My first thought would be to utilize some sort of binning technique like hexagon binning, but I think they're stuck in their desire to visualize it in terms of ellipses.
So I have a non-ideal solution I did in the absence of a proven, more automated technique.
I created a Points layer to create my ellipses, and used layer styles to created ellipses that fit roughly to the areas they indicated on their thumbnail, and with calculations fit the percentages they had estimated. It was all just a series of layer styles and manual selections from there to provide data with the summaries of coverage.
Not a great way in my opinion, but it gets the job done I suppose. I also learned that QGIS has terrible support for creating ellipses, even with plugins.