[Ifeffit] delR and delE

[ritika.uppal at yale.edu]

Shelly, thanks for the suggestions. I'll try these.

Ritika

Quoting "Kelly, Shelly D." <SKelly at anl.gov>:

> Ritika,
> 
> delr and Eo are correlated.  So you can (1) force delr to the same
> for
> two paths with different Eo-values, or (2) use two different
> delr-values
> with one Eo-value.  Most prefer the second choice over the second
> choice.  
> 
> Also, Eo is important at low k-values.  If you move your k-range
> that
> you use in the FT of your data you might find a different Eo value,
> or
> that only one Eo is needed with a shorter k-range than with a
> longer
> k-range.
> 
> I do not think that you model is "okay" with three Eo values.  How
> much
> does EO differ?  Usually Eo is the last button I turn on when I try
> to
> make a model fit the data.  Large Ezero values (> 10 eV) are common
> if
> you have a poor model.  So use only one Eo at the beginning.  Often
> I
> just set it to zero, and then when the model gets close, try to add
> one
> or two more to really get it right on.
> 
> I would try only one Eo first. Let delr vary for all the paths and
> change the k-range.  See how small the k-range needs to be before
> you
> can fit the data with only one E0 value.
> 
> 1.9 Angstroms is a short distance.  It would seem more reasonable to
> me
> that you might need two Eo values.  One for this short shell and
> then
> one more for all the other shells.
> 
> Give these things a try and tell us how it goes.
> 
> Shelly
> 
> > -----Original Message-----
> > From: ritika.uppal at yale.edu [mailto:ritika.uppal at yale.edu] 
> > Sent: Wednesday, October 13, 2004 2:50 PM
> > To: Segre, Carlo U.; XAFS Analysis using Ifeffit
> > Subject: Re: [Ifeffit] delR and delE
> > 
> > 
> > 
> > I have 3 different values of Eo for the fit. 2 for the first 
> > shell and 
> > the 3rd for the rest of the paths. Actually the bond distance 
> > of the 6 
> > oxygens in the first shell differs. 3 of them have a bond length of
> 
> > around 1.9 A and the other 3 around 2.2 A. So using one value of Eo
> 
> > does not fit the second peak very well.
> > 
> > Ritika
> > 
> >  
> > Quoting "Carlo U. Segre" <segre at iit.edu>:
> > 
> > > 
> > > delR is just the deviation from the distance you used in 
> > the feff.inp 
> > > file to generate the paths.  It can be positive or negative 
> > depending 
> > > on
> > > 
> > > whether these atoms are closer or further away from the 
> > central atom.
> > > 
> > > By delE, do you mean Eo?  When you say two values, do you mean 
> > > different values for two different shells?
> > > Carlo
> > > 
> > > 
> > > On Wed, 13 Oct 2004, ritika.uppal at yale.edu wrote:
> > > 
> > > > Hi guys,
> > > > I have a few questions about the fitting parameters. the delR
> for
> > > my
> > > > fit has a negative value. Is that wrong or its possible to have
> a 
> > > > negative value?? and also for the first shell, I have 2 values
> of
> > > del
> > > > E. Does that mean that the model I am using is not a good
> one??
> > > >
> > > > Thanks for all your help
> > > > Ritika
> > > > _______________________________________________
> > > > Ifeffit mailing list
> > > > Ifeffit at millenia.cars.aps.anl.gov 
> > > > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
> > > >
> > > 
> > > --
> > > 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
> > > Carlo.Segre at iit.edu    http://www.iit.edu/~segre
> > > _______________________________________________
> > > Ifeffit mailing list
> > > Ifeffit at millenia.cars.aps.anl.gov
> > > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
> > > 
> > 
> > 
> > _______________________________________________
> > Ifeffit mailing list
> > Ifeffit at millenia.cars.aps.anl.gov
> > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
> > 
> 
> _______________________________________________
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
>