Hello Pei: On Sat, 15 Mar 2008, Richard Mayes wrote:
As for why the feature (low-R junk) increases as Rbkg is adjusted from 1.0 to 1.21 but disappears at 1.22, I'm not sure but my guess is that this is an artifact of removing some of the data that contributes to the feature, and by 1.22 you have removed enough that the feature disappears (hopefully someone will correct me if I'm wrong). I would be weary of using an Rbkg of 1.22 though (as I think you see why in the plots - Question 3).
I think that Rbkg=1.22 is generally a bit too high for comfort is your first shell distances are fairly short. In my opinion, you should set Rbkg below 1.0, in the regions where the low-R peak first disappears. Try changing Rbkg by 0.02 at a time while looking at the k^3 weighting. There is usally a point at which you see a significant reduction in the background peak and then not much more change for a while. The other thing you can do is to model the data with Rbkg=0.9 and 1.22 to see how the fit results compare. This might help you figure out the best way to handle these data. Cheers, Carlo -- Carlo U. Segre -- Professor of Physics Associate Dean for Special Projects, Graduate College Illinois Institute of Technology Voice: 312.567.3498 Fax: 312.567.3494 segre@iit.edu http://www.iit.edu/~segre segre@debian.org