put efficiency in a quadratic function (so the worse efficiency is punished more)
fixed a bug in the previous version where the demand of the subset route is added to the demand of the superset route (this is the wrong way around)
fixed how stations tracked lines in the main algorithm
in routeComplexity, subtracted two stations from total number of stations when calculating station weight, since all paths have at least two stations (doesn't make sense that the origin and destination are factored in the complexity)
some other general weirdness
algorithm now seems to run and give reasonable lines, but i haven't fully ran it yet
changed a few things -