I am working with a student on gerrymandering and he has proposed looking at the polling districts in our county in which the drive time to each polling location is minimized while serving the same approximate population. The districts will be based on the boundaries of the census blocks.
Using the polling places as the facilities and the census block centroids as the demand points, I've tried running a Minimum Impedance solution using the census block population as a weight. From there, I would extract the facilities name and census block FIPS code, dissolving my census blocks based off the facility that my NA output gave me. I've had some solutions, but the populations vary a lot between my solutions.
Any ideas how to solve this? I'm not a NA expert and I'm having a hard time wrapping my head around how to bring in the population / weight component. I know in rural areas, the weight * impedance factor will be higher.
I like the ideas presented in this tool. I've heard of it before. Thank you very much. At the end of the day, it looks like you need to hand select candidate districts based on you input parameters. Maybe I am missing something.
I'm not sure how well it will work in this application because 1) I have the polling places around which I want to create my polygons and 2) I want to create 165 polygons, which would take a long time in this application. If I can combine the 2 where I can optimize a universe variable in the Districting Application, I would be good to go.