Implicit conversion from float -113.10559006211182 to int loses precision