[Ifeffit] Cumulant expansion fittings

[John J. Rehr]

Hi everyone,

  I would like to add to Marco Vaccari's remarks that the physical quantities
involved in the XAFS cumulants involve displacement-displacement correlations,
e.g.  <vec r_i . vec r_j> and hence are sensitive to correlations in lattice
directionality.  Thus for example u_par and u_perp have an implicit pair
of indices

    u_par = (vec r_i - vec r_j). hat R_ij
    vec u_perp = (vec r_i - vec r_j) - u_par hat R_ij

and contain information about the nature of the three dimensional structure
and its fluctuations in a system.

As a consequence these quantities and their ensemble averages also differ
from the corresponding lattice displacements in XRD.

  J. Rehr