Calculating structure factor from radial distribution curves is normal practice but in a publication is it just the real part of the structure factor which is reported or is it the magnitude i.e. (if x+iy is the S(q) corresponding to a particular q then x**2+y**2).