Rounding function

An assignment I'm currently doing required me to round of a float. I didn't know if there was a standard function for it or not, (if there is I couldn't find it) so I wrote my own.

Code:

---

int round(float x){
  int y = (int)x;
  if((x - y) > 0.5f){
    x = (float)++y;
  }
  else{
    x = (float)y;
  }
  return (int)x;
}

---

It works fine but I'm just seeing what people have to say about it. Maybe some tips on improvment.

why are you casting y to a float before assigning it to x? and why are you returning x?

why not just increment or decrement y and return it?

and the answer to your standard function question: kinda. ther's ceil(x) and floor(x). you can probably tell what they do from their names...

edit:

Code:

---

int round(float x)
{
        int y = (int)x;
        if((x - y) > 0.5f)
        {
                ++y;
        }
        return y;
}

---

Test with some negative values too.

Quote:

---

Originally Posted by major_small

why are you casting y to a float before assigning it to x? and why are you returning x?

why not just increment or decrement y and return it?

---

heh, I don't know. I whipped it up pretty quickly. Thanks for the input.

it's work fine

No it doesn't. Read Dave's post.


Quzah.

Well I could test for a negative value and handle it accordingly but my program wont ever use the function with a negative value. If I intended to reuse this function then I would but in this situation it is useless to my program.

Code:

---

#include <math.h>

---

Code:

---

int round(float x) {
    return (int) floor(x + 0.5);
}

---

This should round -1.5 to -1, which is okay depending on what your definition of "up" is. (-1.51 rounds to -2)

Code:

---

int round(float x) {
    return (int) (x - 0.5 + (x > 0.0));
}

---

This should round -1.5 to -2, and 1.5 to 2, which is good if you have this definition of "up".