Square decision region model specification - initial evaluation failed

Thanks, this at least gets me started. A pretty straightforward switch statement decision boundary appears to at least not crash… getting divergences, but at least that’s something to get working off of.