As you can see in the Oracle Documentation for java.awt.Point, x and y are stored as int. However, getX and getY return double. While there is a setLocation method that takes 2 double types, there is no constructor that does. Furthermore, the double gets truncated to an int internally anyway.
Is there a good reason for this? You might avoid a cast on setLocation by having a method that takes double types, but you have the opposite problem when you call getX and getY. There's also a misrepresentation of the precision of the Point class by returning a double from getX and getY.
