[Ifeffit] Why different installations of Atoms generate different results?

Bruce Ravel bravel at bnl.gov
Fri Sep 21 13:36:09 CDT 2012


Ku-Ding,

I am not sure how to respond to this.  To me, it reads like a rather
rambling rant about the things you don't like in my software.  You
haven't really articulated a question.  Nor have you provided any kind
of evidence that Demeter is misbehaving in any way that I can
reproduce on my own computer.

Now, don't get me wrong ... I love a good rant and indulge myself all
the time!  And I certainly agree that there is ample room for
improvement in my codes.  But if you want me to either explain
something or change something, you have to articulate the issue much
more clearly.

That said, I have a few comments that may be helpful....

1. Here is the Artemis Users' Guide:
     http://bruceravel.github.com/demeter/artug/index.html
   It is not complete and I haven't yet written a description of the
   plotting window. 

2. The plotting toolbar in Atoms is intended to provide a crude
   visualization of the paths just calculated.  It is, in no sense,
   intended as a complete visualization tool.  Rather, it's purpose is
   to answer questions like "where in R space is the peak of this
   path?"  or "which of these paths have substantial spectral weight?"
   For such questions it doesn't much matter what k-weight you use.

3. In Artemis, the FT and k-weighting parameters from the Plot window
   are used by the plots from the Atoms/Feff window.  In stand-alone
   atoms, the default values are used.  The configuration tab in the
   stand-alone Atoms does not, however, offer the plot or gnuplot
   configuration groups.  I suppose that could consititute a bug if
   your expectation of those plots exceeds their intent.

4. I no longer support the 0.8 series.  There will be no updates or
   bug fixes forthcoming for those older versions.

B



On Friday, September 21, 2012 11:06:00 PM Tsuei, Ku-Ding wrote:
> Hi Bruce,
> 
> I am a novice of this community and actually have less than two weeks of
> experience using IFEFFIT (version 1.2.11 with iff_013 update) and its
> front end packages Demeter (version 0.9.11) and its predecessor Horae
> (Athena, Artemis etc. version 0.8.014) running under Windows 7
> Professional SP1. They are excellent software guiding me quickly into
> this EXAFS world. Thanks a lot.
> 
> However, I feel confused by viewing different results (chi(k))
> calculated from three installations of Atoms, namely Atoms within
> (D)Artemis, stand alone (D)Atoms, and Atoms within the old Artemis. I
> then used the simplest Au.inp as the input for all. I checked the path
> outputs making sure the generated paths are identical and selected
> individual paths. To my surprise the chi(k) curves have different
> envelopes and magnitudes from 3 Atoms. It seems to me that chi(k) are
> calculated from the same feff6 program and should be identical from any
> installation; no FT range is involved to affect chi(R). Could you help
> clarify my confusion?
> 
> Perhaps the new Demeter can be improved further. It seems that only
> k^2*chi(k) is plotted in Atoms within (D)Artemis, and only k*chi(k) is
> plotted in the stand alone (D)Atoms by viewing the vertical labels in
> plots. There seems no obvious place I can change k-weighting like in the
> old Artemis. I have not been able to find where, such as Edit preference
> in the old Artemis, I can change the default parameters, or learn what
> parameters are used in doing FT (kmin, kmax) or window function and so
> on. I learn these preference parameters quickly owing to the extensive
> documents built-in side by side with "action tabs" in the old Artemis
> and the link to html documents in the old Athena. By the way, watching
> your videos is an excellent way to learn Demeter but I had to pause to
> search related projects or files to run side by side with videos; some
> links with video download sites could be very helpful.
> 
> Look forward to reading your answers and utilizing the more powerful
> Demeter. Thanks again.
> 
> Best regards,
> 
> Ku-Ding Tsuei
> ------------------------
> Ku-Ding Tsuei, PhD
> Associate Scientist
> National Synchrotron Radiation Research Center
> 101 Hsin-Ann Road
> Hsinchu Science Park
> Hsinchu, TAIWAN 30076, ROC
> email: tsuei at nsrrc.org.tw


-- 

 Bruce Ravel  ------------------------------------ bravel at bnl.gov

 National Institute of Standards and Technology
 Synchrotron Methods Group at NSLS --- Beamlines U7A, X24A, X23A2
 Building 535A
 Upton NY, 11973

 Homepage:    http://xafs.org/BruceRavel
 Software:    https://github.com/bruceravel



More information about the Ifeffit mailing list