浮点数比较自己的示例

		float f1 = 0.9999999f; 
		float f3 = 0.99999999f; 
		float f2 = 1f; 
		boolean value = (Float.floatToIntBits(f1) == Float.floatToIntBits(f2)); 
		System.out.println(Float.floatToIntBits(f1));
		System.out.println(Float.floatToIntBits(f2));
		System.out.println(value);
		System.out.println(f1==f2);
		System.out.println("----------");
		boolean value2 = (Float.floatToIntBits(f3) == Float.floatToIntBits(f2));
		System.out.println(f3==f2);
		System.out.println(Float.floatToIntBits(f3));
		System.out.println(value2);

 输出结果:

1065353214
1065353216
false
false
----------
true
1065353216
true

 

可见缺省float的精度是8位小数,即>=8位小数时就认为相等了