格式十进制到没有科学记数法的小数位数

Question

I am currently formatting a double using the code:

myDouble.ToString("g4");

To get the first 4 decimal places. However I find this often switches over to scientific notation if the number is very large or very small. Is there an easy format string in C# to just have the first four decimal places, or zero if it is too small to be represented in that number of places?

For example, I would like:

1000 => 1000
0.1234567 => 0.1235
123456 => 123456 (Note: Not scientific notation)
0.000001234 => 0 (Note: Not scientific notation)

Answer 1

You can try like this:

0.1234567.ToString("0.####")

Also check Custom Numeric Format Strings

#

Replaces the "#" symbol with the corresponding digit if one is present; otherwise, no digit appears in the result string.

Also as Jon as correctly pointed that it will round your number. See the note section

Rounding and Fixed-Point Format Strings

For fixed-point format strings (that is, format strings that do not contain scientific notation format characters), numbers are rounded to as many decimal places as there are digit placeholders to the right of the decimal point.

Answer 2

Use the method.

String.Format("{0:0.####}", 123.4567123); //output: 123.4567

Note: Num of #'s indicate the number of digits after decimal that are required.

Answer 3

I agree with kjbartel comment. I wanted exactly what the original question asked. But his question is slightly ambiguous.

The problem with ### format is it fills the slot if a digit can be represented or not. So it does what the original question asks for some numbers but not others.

My basic need is, and it's a pretty common one, if the number is big I don't need to show decimal places. If the number is small I do want to show decimal places. Basically X number of significant digits.

The "Gn" Format will do significant digits, but it switches to scientific notation if you go over the number of digits. I don't want E notation, ever (same requirement as the question).

So I used fixed format ("Fn") but I calculate the width on the fly based on how "big" the number is.

var myFloatNumber = 123.4567;
var digits = (int) Math.Log10(myFloatNumber);
var maxDecimalplaces = 3;
var format = "F" + Math.Max(0,(maxDecimalplaces - digits));

I swear there was a way to do this in C++ (Visual Studio Flavor) in teh format statement or in C# and perhaps there is, but I can't find it.

So I came up with this. I could have converted to a string and measured length before decimal point as well. But converting it to a string twice felt wrong.