[Problem] Rounding off to the nearest multiple of a variable

Okay, I wanted to know if anyone knows how to create a formula that rounds a number off to the nearest variable. Note, that I am not trying to round the number to the nearest integer but to the nearest mulitple of a variable here are some examples:

code:

---

var staticNum = 42;

var changingNum = 35;

---

now I want to round changingNum to the nearest multiple of staticNum.
In this case I want changingNum to equal 42. How do I do it? Then how would I do it this way:

code:

---

var changingNum = 86;

---

In this case, I want changingNum to equal 84, because 84 is the nearest multiple of 42. Please help. If you don't understand what I mean, ask questions because I will be happy to answer.

Here's a try, I'm asuing you're not alowing negative multples and allowing the use of multiples of 1.

//--Edit little bug found, will add new code in few min's :)

Hope you can understand it, I tried to add comments but I suck at explaining stuff like this :)

thank you so much except there was a little problem with it. I had to change Math.floor to Math.round, because I wanted it to round to the nearest multiple. But thanks! Heres the new script, for anyone who wants it:

code:

---

var startN = 42;

var newN = 158;

if (newN<startN) {

        newN = startN;

} else {

        var multiples = Math.round(newN/startN);

        newN = multiples*startN;

        if (multiples>startN/2) {

                newN += startN;

        }

}

trace(newN);

---

Hey,
I see you've already seen my bugged version of the code ;) I thought it was done, but I didn't test it too well. Here's a working version (atleast I think :D), but I haven't tested it with negative values though. I suggest you use this code as it's easier to read and to adjust for use with negative values.

code:

---

var srtN = 23;

var oldN = 234;

var newN;

//--

if (oldN%srtN<srtN/2) {

	newN = oldN-(oldN%srtN);

} else {

	newN = oldN-(oldN%srtN)+srtN;

}

trace(newN);

---

SaphuA

well if you wanted to make it a negative value can't you just change everything to positvie and then use newN * -1 after its solved to make it a negative number?