Comparing floats using a dynamic tolerance (depending on the size of the fractional part)

Development Programming & Scripting
1

Does UE4 provide functions for float comparisons, e.g. where I can set tolerance/epsilon/ULP?

Usually I would do it by hand, e.g. fabs(f1 – f2) <= epsilon but maybe UE4 already has a better way to support a dynamically tolerance depending on the size of the fractional part of the float?

FMath only seems to support hardcoded tolerances, e.g. FMath::IsNearlyEqual.

2

Thanks for posting, however I’ve edited my question in the same time you posted the answer (my fault. so upvote nevertheless). I’m looking for something that provides a “dynamic” tolerance depending on the size of the fractional part of the float.

3

FMath::IsNearlyEqual(float, float, double);

4

FMath::IsNearlyEqual accepts a tolerance value as the third parameter; it’s not hard coded nor constant. You can determine your tolerance value however you like and pass it in.