Implicit conversion from float -110.24721878862792 to int loses precision