为什么我需要从double转换为int，反之亦然？

Question

Let's say I have a double and an int.

double x = 5;
int y = x;

When I do this, it throws an error because I need to explicitly cast from double to int like this,

int y = (int) x;

for it to work. But when I do

int x = 5;
double y = x;

No error is thrown, and it works perfectly fine. Is there any reasoning behind having to explicitly cast from double to int but not vice versa?

Answer 1

Double is more widened type than int. According to specification you can assign the value in this order:

byte -> short -> int -> long -> float -> double

The main idea:

A widening primitive conversion does not lose information about the overall magnitude of a numeric value.

But:

A narrowing primitive conversion may lose information about the overall magnitude of a numeric value and may also lose precision and range.

Answer 2

Because when you cast an integer to a double it's the same saying 1 or 1.0. But when you cast from double to int it isn't the same saying 1.8 or 1 and then you have to explicitly cast it.

Answer 3

Apart from losing accuracy, they have different sizes.

Think of trying to get a wardrobe out of a truck and into a car boot. That's not going to work, but taking a box from a car boot to a truck, will work.

int is 2 to 4 bytes whereas a double is 8 bytes as explained here TutorialsPoint

Answer 4

According to Java Documentation :

• a int is stored using 32 bits of memory. It represents integers between -2^31 and a maximum value of 2^31-1 .
• a double is stored using 64 bits of memory. It represents floating numbers between 2^-1074 and (2-2^-52)·2^1023 .

int to double

Given the size of a double in memory, it is obvious that a double can contain an int (who can do more can do less).

double to int

This cast is not permitted since we lose information. If we would like to convert for example a floating number like 1.1618, we lose information. "Equivalents" of this double could be 1 or 2 but Java will never decide for you between the possibilities through the symbol '='. You have to use a function like for example Math.round(1.1618); to choose which conversion you want to transform your double to an int.