On 10/30/2013 11:40 AM, Georges Siddiqi wrote:
Regarding your suggestions, I'm a bit unclear how I can implement this
in Artemis. Is this approach similar to one of your papers:
where you relate the nanoparticle coordination number to the bulk
coordination number via particle radius? (equation 4)
How can I do this in Artemis when there's other variables such as
deltaE, debye-waller factor and delr to worry about?
This is the central problem of EXAFS analysis, or indeed of any
non-linear analysis. Artemis reports on uncertainties, correlations,
and reduced chi-square specifically so you can figure these things
out. Somehow, you have to evaluate whether your fitting model is
defensible and whether the results you are getting from your fits are
defensible. This has to be done ins the presence of correlations.
I talk about this some here:
I yap on and on using that presentation in the "Fit Evaluation and
Fitting multiple structures" talk at
Scott's discussion of these issues in his new book is quite helpful.
Bruce Ravel ------------------------------------ bra...@bnl.gov
National Institute of Standards and Technology
Synchrotron Science Group at NSLS --- Beamlines U7A, X24A, X23A2
Upton NY, 11973
Ifeffit mailing list