Implicit conversion from float -110.098814229249 to int loses precision