Implicit conversion from float -3.296703296703299 to int loses precision