qRound problem!

[SkripT]

Hi all, this time i have a little problem with qRound. Based on the tutorial:

int qRound ( double value )
Rounds value up to the nearest integer.

So, I think that the result of, for example, 256*2.22 (=568.32) should be 569 but the result that qRound gives me is 568. In my point of view this is not rounded up to the next integer. Anybody could explain me why? Thanks

[jacek]

It looks like that "up" shouldn't be there.

inline int qRound(qreal d)
{ return d >= 0.0 ? int(d + 0.5) : int(d - int(d-1) + 0.5) + int(d-1); }

inline int qRound(qreal d)
{ return d >= 0.0 ? int(d + 0.5) : int(d - int(d-1) + 0.5) + int(d-1); }

To copy to clipboard, switch view to plain text mode 

[GreyGeek]

Hi all, this time i have a little problem with qRound. Based on the tutorial:

So, I think that the result of, for example, 256*2.22 (=568.32) should be 569 but the result that qRound gives me is 568. In my point of view this is not rounded up to the next integer. Anybody could explain me why? Thanks

I ran into similar problems and discovered, after searching the topic in Google, that there is an entire sub-branch of C++ devoted to rounding algorithms. Someone offered this:

round(x) = floor( x + 0.5f);

round(x) = floor( x + 0.5f);

To copy to clipboard, switch view to plain text mode 

I borrowed this method from one of the many offerings:

float homestead::Round(float Value, int NumPlaces) {
	int k;
	float Factor;
	float Temp;
	Factor = 1.0;
	for (k = 0; k < NumPlaces; k++)
		Factor = Factor * 10.0;
 
	Temp = Value * Factor + 0.5;
	return Temp / Factor;
}

float homestead::Round(float Value, int NumPlaces) {
	int k;
	float Factor;
	float Temp;
	Factor = 1.0;
	for (k = 0; k < NumPlaces; k++)
		Factor = Factor * 10.0;

	Temp = Value * Factor + 0.5;
	return Temp / Factor;
}

To copy to clipboard, switch view to plain text mode 

because it allows me to specifiy the number of roundoff places.

[SkripT]

Hi thanks both. Finally what I've done is:

int roundUp(double v)
int aux = (int) v;
return v - aux > 0 ? aux + 1 : aux;

int roundUp(double v)
int aux = (int) v;
return v - aux > 0 ? aux + 1 : aux;

To copy to clipboard, switch view to plain text mode 

[yop]

If I understand correctly you are looking for http://www.cppreference.com/stdmath/ceil.html

[dimitri]

This is indeed a bug either in the documentation or the Qt code. It will be fixed.

[wysota]

As far as I understand "round up" is an english expression (phrasal verb) which doesn't mean "round upwards" (just like "look up sth in the dictionary" doesn't tell you to look in the sky right?), but it can be misleading

[jacek]

As far as I understand "round up" is an english expression (phrasal verb) which doesn't mean "round upwards" (just like "look up sth in the dictionary" doesn't tell you to look in the sky right?)

My dictionary says that when it comes to maths "to round up" means "to round upwards", in other contexts it behaves like you wrote.

but it can be misleading

Documentation shouldn't be misleading in any way.

[wysota]

Documentation shouldn't be misleading in any way.

Sure, I don't say it should.

Edit: and I found "round up" in math meaning in my dictionary too