Implicit conversion from float -5.608108108108112 to int loses precision