iphabs nabs rclabs

A ``configuration'' average over the spectra of multiple absorbing atoms is done if the CFAVERAGE card is used. CFAVERAGE currently assumes phase transferability, which is usually good for EXAFS calculations, but may not be accurate for XANES. Note that the CFAVERAGE card is currently unreliable in general, and in particular is incompatible with the DEBYE card for options other than the correlated Debye model ( $\mathtt{idwopt} > 0$).

potential index for the type of absorbing atoms over which to make the configuration average (any potential index is allowed).

the configuration average is made over the first nabs absorbers in the `feff.inp' file of type iphabs. You do not need to have potential of index 0 in your input file when using the CFAVERAGE card, but you must have the same type of potential for iph=0 and iph=iphabs. The configurational average is done over ALL atoms of type iphabs, if nabs is less than or equal to zero.

radius to make a small atom list from a bigger one allowed in `feff.inp'. Currently the parameter controlling the maximum size of the list, natxx, is set to 100,000, but this can be increased. The pathfinder will choke on too big an atoms list. You must choose rclabs to have fewer than 1,000 atoms in the small atom list. If your cluster is less 1,000 atoms simply use rclabs=0 or negative always to include all atoms.
Default values are iphabs=0, nabs=1, rclabs=0 (where $\mathtt{rclabs}=0$ means to consider an infinite cluster size).
  *average over all atoms with iph=2 in feff.inp with less than 1000 atoms
  CFAVERAGE 2  0  0