Implicit conversion from float -21.658031088082907 to int loses precision