Implicit conversion from float -110.93457943925233 to int loses precision