Implicit conversion from float -110.41420118343194 to int loses precision