Compare generated models with frequency analysis

[davepagurek]

The goal here is to find a way of measuring the consistency of models we generate. They should have similar general shapes but different details.

Some papers discussing this:

• Using DFT: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.71.3013&rep=rep1&type=pdf
• Using spherical harmonics: https://pdfs.semanticscholar.org/da27/5dcffd835ddfd41fd73ea147c767c605d3f0.pdf

We probably want to compare multiple models made with the same guiding vectors and multiple models made with different ones to ensure that we can see a discernible difference (plus maybe also compare models made using other methods?)