Float UnderFlow Error

[ZonKY420]

so I’m not that experienced with c++ yet.
I believe that the problem is that there’s a float underflow error.
but I’m not really sure how to go about solving that.

would really appreciate some help on this

// Adjust the starting position for the sphere sweep to match the bottom of the capsule.
FVector SweepStart = CapsuleLocation - (CharacterUpVector*(PawnHalfHeight - PawnRadius));

SweepStart  {X=5.490716046094e-315#DEN Y=0.0000000000000000 Z=-515.55547790540504} UE::Math::TVector<double>

CapsuleLocation {X=-515.55644955580078 Y=3297.8822023619878 Z=97.441433419731169} const UE::Math::TVector<double> &

CharacterUpVector {X=0.00097976236391282383 Y=-0.72453746722546297 Z=0.68923471956382287}   UE::Math::TVector<double>

PawnHalfHeight    55.0000000  float

PawnRadius   35.0000000   float

also when crouching so with a smaller pawnhalfheight the float underflow error goes away

[ZonKY420]

yh so i simplyfied the problem more.

FVector tempvalue = FVector(0, 0, 55);
	
FVector SweepCapsuleCenter = FVector(CapsuleLocation.X - tempvalue.X, CapsuleLocation.Y - tempvalue.Y, CapsuleLocation.Z - tempvalue.Z);

so for some reason even if im only subtracting 55 from just the z axis it make the other axis also change/break

{X=1.260373362460e-311#DEN Y=6.951683486879e-310#DEN Z=2.337630270699e-311#DEN}```
which does not make any sense to me

[ZonKY420]

so if you look at a value in breakpoints u need to make sure that the value actually has been calculated before taking the value as proof that its not your code thats bad but the program. because if i just waited for it to be calculated it would have been the correct value