How can you get the number of digits contained in a double?

Question

I'm trying to get the number of digits in the following double value: 56.46855976 without using converting it to a string (and simply replacing the "." with a "").

Anybody got any ideas?

Aww shucks, what is wrong with: `double d = 56.46855976; int length = d.ToString().Replace(".", "").Length;` — slugster, Sep 16 '10 at 11:50
Take into account also that the number you see (56.46855976) is only a *representation* of a value. The value 1/3 is very short, the number 0.33333... isn't. So except for view issues there is not much information in the length of a number. — Martin Hennings, Sep 16 '10 at 11:54
Possible duplicate of [c# Decimal to string for currency](https://stackoverflow.com/questions/10437416/c-sharp-decimal-to-string-for-currency) — TylerH, Feb 16 '18 at 16:59

score 5 · Answer 1 · answered Sep 16 '10 at 11:48

Count how often you must divide the number by 10 until it's smaller than 1 -> that gives you the digits before the decimal point.

Then count how often you must multiply the original number by 10 until it equals the Math.Floor-result -> that gives you the digits behind the decimal points.

Add. Be glad.

Edit: As Joey points out, there is some uncertianity in it. Define a maximum number of digits beforehand so you don't create an infinite loop. On the other hand - "How long is the coast of Denmark?"...

score 1 · Answer 2 · answered May 31 '17 at 11:09

    /// Returns how many digits there are to the left of the .
    public static int CountDigits(double num) {
        int digits = 0;
        while (num >= 1) {
            digits++;
            num /= 10;
        }
        return digits;
    }

As Martin mentioned, counting to the right of the . is pointless.

Tests:

MathPlus.CountDigits(56.46855976)                 -> 2
MathPlus.CountDigits((double)long.MaxValue + 1)   -> 19
MathPlus.CountDigits(double.MaxValue)             -> 309

score 0 · Answer 3 · answered Sep 16 '10 at 11:48

Converting to a string might be the best option you have. Remember that doubles are represented in Base 2 internally. Therefore the decimal representation you see is only an approximation of the actually stored value (except for integers up to 2⁵³) which is a sum of individual powers of 2.

So trying to figure out the number of decimal digits from the binary representation is certainly not an easy or trivial task – especially since the framework might also apply rounding to make numbers like 3.999999999998 appear like 4.0 since they appear to have more precision than there actually is.

How can you get the number of digits contained in a double?

3 Answers3

Linked