Victor raised a very good point. To extend his suggestion, I would find the best fit parameters by fitting the original data. Then, I would fit the back-Fourier-transformed data in the same exactly r-range as before, using the same q-range as the k-range used in the first fit. All fit parameters should be fixed to the best fit values found earlier and run the "fit" again (nothing is going to be varied). This will assure that you are varying only one factor in this comparison: the spectral content. Anatoly -----Original Message----- From: ifeffit-bounces@millenia.cars.aps.anl.gov on behalf of vikrai@comcast.net Sent: Wed 7/6/2005 12:22 PM To: XAFS Analysis using Ifeffit Cc: Subject: Re: [Ifeffit] R-factor Alison, I don't think that the difference of R-factors you mention is a big deal. You did not mention in your message more important thing. Did you get the same values of parameters? Besides, it is hardly to expect to find a signal caused by photoelectron backscattering at R<1 A. So the lower limit in R-space fitting equal to 0.1 A is too small probably. As well as the lower limit in q-space fitting because in proteins in this momentum region only full multiple scattering approach should work correct. Victor Krayzman -------------- Original message --------------

Bruce,

I've looked, and my R-factors for the back-transformed space are not necessarily twice the value of the R-factors for the k-space fits. In fact, they are often quite close. For example, just recently I had a k-space R-factor of 0.102 and a q-space R-factor of 0.113. Now, I realize those numbers are very close, but I'm afraid if I try to publish this, then I will get criticism for the q-space R-factors being larger. If I can explain it, then maybe it won't be a problem.

I do see what you're saying about the complex vs. real function, and that definately makes sense, but they don't differ by a factor of 2. Hmmm.

alison

On Wed, 6 Jul 2005 09:43:03 -0500 Bruc! e Ravel wrote:

...

On Tuesday 05 July 2005 18:39, Alison Costello wrote:

...

I have a question regarding the calculation of the R-factor. I am currently using SIXPACK to process my protein EXAFS data, and I fit the data in both k-space and in q-space (back-transformed k-space). I typically will use k-ranges = [1-13.6] and for multiple-scattering fits, will use an R-range of 0.1 - 4.5. Sometimes, the R-factor for the q-space fit is greater than the R-factor for the k-space fit, which should not be true, as the back-transformed space filters out noise. I am confused as to why this occurs, and am wondering if the R-factor for the back-transformed space is calculated differently than for k-space f! its.

A lison,

I would expect the R factor to be bigger in k space due to the high frequency portions of the data that get filtered out of the q space data. That is what happened in the one example I just tried.

What are some example values of the R-factors? Do they differ by about a factor of 2?

One thing that occurs to me is that chi(q) is a complex function while chi(k) is a real function. In Ifeffit those two functions should have the same number of points, but the R factor in q is computed using both parts of the complex function. Thus it would seem reasonable if they differed by about 2 and there were no significant Fourier components in the data beyond your rmax.

B > > -- & gt; > Bruce Ravel ----------------------------------- bravel@anl.gov -or- ravel@phys.washington.edu Environmental Research Division, Building 203, Room E-165 Argonne National Laboratory phone: (1) 630 252 5033 Argonne IL 60439, USA fax: (1) 630 252 9793

My homepage: http://feff.phys.washington.edu/~ravel EXAFS software: http://feff.phys.washington.edu/~ravel/software/exafs/

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit