为什么使用整数而不是双精度？

Question

I have code for two versions of the same function: one iterative and one recursive. I want to know why the return type is int . What does it afford you? Why can't the return type be double ?

public static int iterative(int n){
  int var=1;
  int var1=1;
  int var2=1;

  if (n<=1) {
    return 1;
  }

  for(int i=2;i<=n;i++){
    var=var1+var2+1;
    var2=var1;
    var1=var;
  }

  return var;
}

private static int recursive(int n){
  if (n<=1){
    return 1;
  }

  return rec(n-1)+rec(n-2)+1;
}

public static void main(String[] args) {
  System.out.println(iter(10));
}

Answer 1

You could certainly define the function to return a double , in the same way you could define it to return a String . The formats are capable of representing the values you'll be returning, there's just no reason to do it.

• The function you're computing (a modified Fibonacci sequence) is defined over the set of integers - in fact, only positive integers. It will never produce a value that can not be represented by an int .
• double (and String ) is more expensive in both space and computation.
• The return type is a strong indication to other developers what kind of data to expect. Returning a double that will only contain 1, 2, 3, ... is to some extent misleading.

So use double if you like; it will work just fine. But the question is more "why?" than "why not?".