wrong result.

[Jacobhaha]

In my program i have the following: 84.74 - 84 - 0.74 and it says it equals to -2.14577e-006 , how do i fix this? shouldn't it be 0?

i have:
int = 84 ,
double = 84.74 ,
float = 0.74

[TheBadCoder]

post the code in brackets please...

Let me take a look at it...

[Zhuge]

Floating point representations are not exact, and -2.14577×10^-6 is basically 0.
https://en.wikipedia.org/wiki/Floating_point#Accuracy_problems

[Jacobhaha]

i'd rather not post it all, i will have to change alot of it, but what i posted above is all included in my problem.

[TheBadCoder]

[quote = Zhuge]Floating point representations are not eact[/quote]

You beat me to it .-.

[Jacobhaha]

what do you suggest i do?

[Zhuge]

Instead of expecting exact results from floating point math, handle things based on a range. For example, instead of checking to see if that your computation returned exactly 0, check if it is, say, in the range of [-0.01, 0.01] (or whatever range is appropriate for your work).

[Jacobhaha]

and another stupid question, how would you suggest i do this, with switch?

[Ganado]

Well first off, why are you intermixing doubles and floats? It might not still be exactly zero, but I'm pretty sure it would be at least closer if you did a double minus a double instead of a double minus a float.

Anyway, you'd want to do something like this

const double Threshold = 0.0001;
if ( abs( my_value - what_my_value_should_be ) < Threshold)
{
    cout << "It's close enough to zero.\n";
}