I'm no good with function nodes being very new to TG, but it seems as though the math wouldn't be that hard if you could just orient a couple different parametric equations, granted the inner workings of the function nodes could make this quite an undertaking.
