Implicit conversion from float -110.58139534883722 to int loses precision