Implicit conversion from float -11.422764227642276 to int loses precision