If g(x) is an inverse of f(x), then how do the ordinate and abscissa coordinates of the points on f(x) map onto coordinates of g(x)?