Implicit conversion from float -1.621621621621614 to int loses precision