One has to use the geometrical stade for the theoretical model. This gives better results. Eratosthenes' approximation has a 6.5% error with the same Google data. The relative error shouldn't depend on the units that are used. It was the model that was off.
No comments:
Post a Comment