Implicit conversion from float -175.43778801843314 to int loses precision