Implicit conversion from float -79.29133858267716 to int loses precision