Indeed, numerical math was only my second and third semester... =)
(maybe technical universities are different... )
For me worst case of error propagation is reason enough to avoid floating point calculations for "disguised" integers, whenever possible.
And that's nearly always the case if people start to worry about "strange errors", e.g. when calculating cents as fractals fractions ¹...
(1) for JadeNB an amusing typo! 8) |