Is it possible convert `Int.MAX_VALUE` to a `Double`? I’ve tried `toDouble()`, but that outputs 2.14...

# getting-started

[Shalom Halbert]

Is it possible convert Int.MAX_VALUE to a Double? I’ve tried toDouble(), but that outputs 2.147483647E9

[ephemient]

Int.MAX_VALUE.toDouble() is exact, you're just seeing Double.toString() truncating to 9 digits of precision by default

[ephemient]

"%.10g".format(Int.MAX_VALUE.toDouble())

will show 10 digits of precision, which is enough for all digits of Int.MAX_VALUE

[Shalom Halbert]

So the reason Int.MAX_VALUE.toDouble() equals 2.147483647E9 instead of 2147483647.0 is because that requires 11 digits? That’s confusing because the Kotlin docs say there can be “15-16” decimal digits. To clarify the situation, below are the cases referenced so far.

Int.MAX_VALUE // 2147483647
Int.MAX_VALUE.toDouble() // 2.147483647E9
"%.10g".format(Int.MAX_VALUE.toDouble()) // "2147483647"

[ephemient]

Int.MAX_VALUE.toDouble() does not equal 2.147483647E9, that is just its string representation

[ephemient]

most Doubles have string representations that do not parse back to the same Double

[Joffrey]

equals 2.147483647E9 instead of 2147483647.0

@Shalom Halbert both of these are the same number. The first one is simply written in scientific notation (with E9 at the end meaning x10^9), while the other is in standard notation. It's just a matter of how you print the string representation of the number, as @ephemient pointed out.

What is your issue with this output exactly? Is it because of the scientific notation, or is it because you believe the value is not the same?

[ephemient]

actually I was mistaken to a certain extent; it is supposed to emit the shortest string that can uniquely reproduce the same Double. but (at least on JVM) it's buggy: https://bugs.openjdk.java.net/browse/JDK-4511638

[ephemient]

in any case you're not running into that here, it's simply a different representation than you may be expecting, but it is still the same Double

[Shalom Halbert]

You folks have it correct, I didn't realize that toDouble() converts the Int into the same number, using scientific notation. Thank you for helping me realize that I was reading it wrong

[Joffrey]

Just to be 100% clear, the conversion done by Int.toDouble() is format-agnostic, it converts the int value into the same value of type Double. The scientific notation only matters (or even exists) when you try to print the number

[ephemient]

int.toDouble().toInt() == int // for all Int, but not all Long
double.toString().toDouble() == double // except NaN

[Joffrey]

I meant Int.toDouble() , of course when strings are involved it's all about format