Consume Product

It's called PCA. Principle component analysis. It's a form machine learning where you input all N- dimensional information related to some labels and the computer looks through all the data to separate them into groups based on these labels. The computer tries to find a balance between distinctive groups and shared values between groups. The X and Y axis are some incomprehensible heuristic of information to generate a comprehensible image. Because genetics are N-dimensional (literally thousands and thousands of dimensions), this is really the only way to comprehend distinctions between groups on a large scale.

https://en.wikipedia.org/wiki/Principal_component_analysis