|Status:||Backlog||Start date:||28 Jul 2016|
Is there need for a tolerance_factor? Probably not.
Are small numbers meant on a linear or a logarithmic scale? Is 1e-100 - 0 a small relative difference?
Why isn't get_relative_difference(a,b) symmetric in a<->b?
What are the consequences for simulations and fits?
By the way: areAlmostEqual is used only once in the new Fit/ directory, and should therefore be replaced by a local routine, so that Numeric.h can be moved to Core/.