- Ifeffit - millenia.cars.aps.anl.gov

Re: [Ifeffit] question about Eo and simultaneous fitting
by Kelly, Shelly D. 14 Feb '07

14 Feb '07

Hi Scott, The change in frequency is a bigger affect at high k than the energy shift parameter. The energy shift parameter affects the data more strongly at low k. So I align the data to each other at low k, and then they become out of phase with each other at high k. A similar effect is seen for pressure dependent data where the oxidation state doesn't change but the bond length changes. I also look at how well the data are both aligned to the theory used to describe the data sets independently. The energy shift values should overlap within the uncertainty to justify using a single parameter for them. I suppose that my method might not work for all systems. But if the data are similar enough so that a multiple data set fit has significant overlap in the models used to describe the spectra, then it has worked. Shelly > -----Original Message----- > From: ifeffit-bounces(a)millenia.cars.aps.anl.gov [mailto:ifeffit- > bounces(a)millenia.cars.aps.anl.gov] On Behalf Of Scott Calvin > Sent: Wednesday, February 14, 2007 9:55 AM > To: XAFS Analysis using Ifeffit > Subject: Re: [Ifeffit] question about Eo and simultaneous fitting > > Hi Shelly, > > Interesting--I've never tried that method with samples that I expect > to differ in oxidation state. If it's differing in oxidation state, > the local environment is probably different too...particularly, > nearest-neighbor distances should be changed significantly. So how do > you align the chi(k) data with each other, when they oscillate at a > different frequency? > > --Scott Calvin > Sarah Lawrence College > > At 10:38 AM 2/14/2007, you wrote: > > >The approach that I like to take, is to vary the choice of E0 in the > >background subtraction step so that the chi(k) spectra are well aligned > >with each other and the theory. Then only one Ezero parameter is > >needed. > > > >Cheers, > >Shelly > > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

1 0

Re: [Ifeffit] question about Eo and simultaneous fitting
by Kelly, Shelly D. 14 Feb '07

14 Feb '07

Hi Tony, The approach that I like to take, is to vary the choice of E0 in the background subtraction step so that the chi(k) spectra are well aligned with each other and the theory. Then only one Ezero parameter is needed. Cheers, Shelly > -----Original Message----- > From: ifeffit-bounces(a)millenia.cars.aps.anl.gov [mailto:ifeffit- > bounces(a)millenia.cars.aps.anl.gov] On Behalf Of Scott Calvin > Sent: Wednesday, February 14, 2007 7:52 AM > To: XAFS Analysis using Ifeffit > Subject: Re: [Ifeffit] question about Eo and simultaneous fitting > > Hi Tony, > > A quick answer before I try to fight my way through an ice storm to > get to work. :) > > Fitting the three spectra simultaneously makes sense, but if > oxidation state is expected to be different, then you really need > different E0's. > > As far as doping, it depends on the details. In some cases, it may > require a different E0's, in others, not. Of course, you can try it > both ways, and see what the fits think. > > --Scott Calvin > Sarah Lawrence College > > At 05:04 AM 2/14/2007, you wrote: > >Hi, > > > >first I would like to thank all experienced people answering in > >this forum!! I have learned really a lot from you. > > > >And now my question: > >Should I fit EXAFS spectra of a doped material and the same > >material, but oxidized and reduced (three spectra) simultaneously? > >In my opinion: yes, as I am looking for small differences. Is > >varying one Eo for the three samples the right approach? > >I am also wondering, should I vary one Eo for samples with different > >doping concentrations? > > > > > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

1 0

Debian packages
by Carlo Segre 13 Feb '07

13 Feb '07

Hello All: I have set up a new repository for the Debian packages of IFEFFIT & friends. I have also begun to build Ubuntu packages as well. Of course, whenever possible the packages are in the Debian archive. The following are available to use in your /etc/apt/sources.list # backported packages for sarge (will stop supporting this once etch # releases) deb http://debian-xray.iit.edu sarge main contrib non-free # backported packages for etch (now that etch is in freeze, any new # versions will be placed here) deb http://debian-xray.iit.edu etch main contrib non-free # additional software not yet in Debian (you can see what is there just # for grins!) deb http://debian-xray.iit.edu sid main contrib non-free # packages made for Ubuntu (I will support only the last three releases) deb http://debian-xray.iit.edu dapper main contrib non-free deb http://debian-xray.iit.edu edgy main contrib non-free deb http://debian-xray.iit.edu feisty main contrib non-free The old repositories at fermi.phys.iit.edu/debian are still there but not being updated. I will probably remove them in 2007. 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(a)iit.edu http://www.iit.edu/~segre segre(a)debian.org

1 0

feff and xanes calculation troubles
by Paul Fons 13 Feb '07

13 Feb '07

Hello all, I am working with a colleague who is working on some High-Tc materials. He has data taken at the F K-edge and is trying to simulate these spectra using feff8 (the same problem occurs from feff8.1 to feff8.5). To make a short story short, the SCF routine never terminates (converges). A bit of the output is attached below. The problem is a little nasty in that while the data is being taken from the F K-edge, the material contains heavy atoms (Ba). I am pretty much at a loss of what to try to improve the situation. I have attempted to change the augmentation sphere overlap using both the AFOLP and FOLP cards to no avail (same problem). I have also tried to use smaller steps and change the amount of mixing in the initial SCF step (using the options on the SCF card). I am running out of ideas of what to try next and was wondering if anyone had any suggestions. The SCF radius is set to the second coordination shell, but moving it around doesn't help with the lack of convergence. The calculation does run to completion without the SCF step, but this leaves a lot to be desired. Thanks for any help out there. I submitted this to the feff mailing list last week, but all is quiet, so I am trying here (sorry for the duplication if that counts). Paul Number of processors = 4 Feff 8.50 XANES: 0234F2.0 (0234f_j2) Executing pot on 4 Number of processors = 4 Calculating potentials ... free atom potential and density for atom type 0 free atom potential and density for atom type 1 free atom potential and density for atom type 2 free atom potential and density for atom type 3 free atom potential and density for atom type 4 free atom potential and density for atom type 5 initial state energy overlapped potential and density for unique potential 0 overlapped potential and density for unique potential 1 overlapped potential and density for unique potential 2 overlapped potential and density for unique potential 3 atomic xmu,vint,rhoint -0.209299551987679 -0.698008117675781 0.410117955794590 atomic xmu,vint,rhoint -0.209299551987679 -0.698008117675781 0.410117955794590 atomic xmu,vint,rhoint -0.209299551987679 -0.698008117675781 0.410117955794590 overlapped potential and density for unique potential 4 overlapped potential and density for unique potential 5 muffin tin radii and interstitial parameters atomic xmu,vint,rhoint -0.209299551987679 -0.698008117675781 0.410117955794590 iph, rnrm(iph)*bohr, rmt(iph)*bohr, folp(iph) 0 1.10153E+00 1.07171E+00 1.06941E+00 1 1.97019E+00 1.91470E+00 1.07252E+00 2 1.69963E+00 1.64348E+00 1.08663E+00 3 1.27604E+00 1.18970E+00 1.15000E+00 4 1.10566E+00 1.05671E+00 1.12119E+00 5 1.14759E+00 1.11204E+00 1.08060E+00 Core-valence separation mu_old= -3.671 SCF ITERATION NUMBER 1 OUT OF 30 Calculating energy and space dependent l-DOS. It takes time ... point # 1 energy = -42.000 Doing FMS for a cluster of 19 atoms around iph = 0 0 FMS matrix (LUD) at point 1, number of state kets = 106 Doing FMS for a cluster of 19 atoms around iph = 1 0 FMS matrix (LUD) at point 1, number of state kets = 126 Doing FMS for a cluster of 23 atoms around iph = 2 0 FMS matrix (LUD) at point 1, number of state kets = 167 Doing FMS for a cluster of 23 atoms around iph = 3 0 FMS matrix (LUD) at point 1, number of state kets = 162 Doing FMS for a cluster of 20 atoms around iph = 4 0 FMS matrix (LUD) at point 1, number of state kets = 120 Doing FMS for a cluster of 19 atoms around iph = 5 0 FMS matrix (LUD) at point 1, number of state kets = 106 point # 20 energy = -30.938 point # 40 energy = -10.825 point # 60 energy = -6.803 point # 80 energy = -6.553 Electronic configuration iph il N_el 0 0 2.047 0 1 5.690 0 2 0.000 0 3 0.000 0 4 0.000 1 0 2.330 1 1 6.314 1 2 2.163 1 3 0.000 1 4 0.000 2 0 0.370 2 1 6.565 2 2 2.462 2 3 0.000 2 4 0.000 3 0 0.489 3 1 0.602 3 2 6.195 3 3 0.000 3 4 0.000 4 0 1.898 4 1 5.097 4 2 0.000 4 3 0.000 4 4 0.000 5 0 2.052 5 1 5.493 5 2 0.000 5 3 0.000 5 4 0.000 mu_new= -6.564 Charge transfer: iph charge(iph) 0 0.053 1 -0.161 2 -0.280 3 0.741 4 -0.198 5 -0.108 SCF ITERATION NUMBER 2 OUT OF 30 Calculating energy and space dependent l-DOS. It takes time ... point # 1 energy = -42.000 Doing FMS for a cluster of 19 atoms around iph = 0 0 FMS matrix (LUD) at point 1, number of state kets = 106 Doing FMS for a cluster of 19 atoms around iph = 1 0 FMS matrix (LUD) at point 1, number of state kets = 126 Doing FMS for a cluster of 23 atoms around iph = 2 0 FMS matrix (LUD) at point 1, number of state kets = 167 Doing FMS for a cluster of 23 atoms around iph = 3 0 FMS matrix (LUD) at point 1, number of state kets = 162 Doing FMS for a cluster of 20 atoms around iph = 4 0 FMS matrix (LUD) at point 1, number of state kets = 120 Doing FMS for a cluster of 19 atoms around iph = 5 0 FMS matrix (LUD) at point 1, number of state kets = 106 point # 20 energy = -30.863 point # 40 energy = -10.614 point # 60 energy = -6.564 point # 80 energy = -6.814 point # 100 energy = -7.064 point # 120 energy = -7.314 point # 140 energy = -7.564 point # 160 energy = -7.814 point # 180 energy = -8.064 point # 200 energy = -8.314 point # 220 energy = -8.564 point # 240 energy = -8.814 point # 260 energy = -9.064 point # 280 energy = -9.314 point # 300 energy = -9.564 point # 320 energy = -9.814 point # 340 energy = -10.064 point # 360 energy = -10.314 point # 380 energy = -10.564 point # 400 energy = -10.814 point # 420 energy = -11.064 point # 440 energy = -11.314 point # 460 energy = -11.564 point # 480 energy = -11.814 point # 500 energy = -12.064 Electronic configuration iph il N_el 0 0 1.953 0 1 5.525 0 2 0.000 0 3 0.000 0 4 0.000 1 0 2.185 1 1 6.138 1 2 0.633 1 3 0.000 1 4 0.000 2 0 0.296 2 1 6.452 2 2 0.726 2 3 0.000 2 4 0.000 3 0 0.511 3 1 0.579 3 2 10.062 3 3 0.000 3 4 0.000 4 0 1.853 4 1 4.537 4 2 0.000 4 3 0.000 4 4 0.000 5 0 1.941 5 1 5.032 5 2 0.000 5 3 0.000 5 4 0.000 mu_new= -12.053 Charge transfer: iph charge(iph) 0 0.137 1 0.054 2 -0.136 3 0.582 4 -0.233 5 -0.084 SCF ITERATION NUMBER 3 OUT OF 30 Calculating energy and space dependent l-DOS. It takes time ... point # 1 energy = -42.000 Doing FMS for a cluster of 19 atoms around iph = 0 0 FMS matrix (LUD) at point 1, number of state kets = 106 Doing FMS for a cluster of 19 atoms around iph = 1 0 FMS matrix (LUD) at point 1, number of state kets = 126 Doing FMS for a cluster of 23 atoms around iph = 2 0 FMS matrix (LUD) at point 1, number of state kets = 167 Doing FMS for a cluster of 23 atoms around iph = 3 0 FMS matrix (LUD) at point 1, number of state kets = 162 Doing FMS for a cluster of 20 atoms around iph = 4 0 FMS matrix (LUD) at point 1, number of state kets = 120 Doing FMS for a cluster of 19 atoms around iph = 5 0 FMS matrix (LUD) at point 1, number of state kets = 106 point # 20 energy = -31.019 point # 40 energy = -12.053 point # 60 energy = -11.990 point # 80 energy = -11.740 point # 100 energy = -11.490 point # 120 energy = -11.240 point # 140 energy = -10.990 point # 160 energy = -10.740 Electronic configuration iph il N_el 0 0 1.960 0 1 5.591 0 2 0.000 0 3 0.000 0 4 0.000 1 0 2.194 1 1 6.168 1 2 0.690 1 3 0.000 1 4 0.000 2 0 0.311 2 1 6.485 2 2 0.838 2 3 0.000 2 4 0.000 3 0 0.493 3 1 0.564 3 2 9.770 3 3 0.000 3 4 0.000 4 0 1.858 4 1 4.581 4 2 0.000 4 3 0.000 4 4 0.000 5 0 1.947 5 1 5.142 5 2 0.000 5 3 0.000 5 4 0.000 mu_new= -10.658 Charge transfer: iph charge(iph) 0 0.419 1 0.862 2 0.317 3 0.212 4 -0.419 5 -0.088 SCF ITERATION NUMBER 4 OUT OF 30 Calculating energy and space dependent l-DOS. It takes time ... point # 1 energy = -42.000 Doing FMS for a cluster of 19 atoms around iph = 0 0 FMS matrix (LUD) at point 1, number of state kets = 106 Doing FMS for a cluster of 19 atoms around iph = 1 0 FMS matrix (LUD) at point 1, number of state kets = 126 Doing FMS for a cluster of 23 atoms around iph = 2 0 FMS matrix (LUD) at point 1, number of state kets = 167 Doing FMS for a cluster of 23 atoms around iph = 3 0 FMS matrix (LUD) at point 1, number of state kets = 162 Doing FMS for a cluster of 20 atoms around iph = 4 0 FMS matrix (LUD) at point 1, number of state kets = 120 Doing FMS for a cluster of 19 atoms around iph = 5 0 FMS matrix (LUD) at point 1, number of state kets = 106 point # 20 energy = -30.879 point # 40 energy = -10.658 point # 60 energy = -10.608 point # 80 energy = -10.358 dot ... dot ... dot point # **** energy = 371.142 point # **** energy = 371.392 point # **** energy = 371.642 point # **** energy = 371.892 point # **** energy = 372.142 point # **** energy = 372.392 point # **** energy = 372.642 point # **** energy = 372.892 point # **** energy = 373.142 crash! bang!

1 0

Re: [Ifeffit] ERROR in getting feff8 input file for YBCO from Web atoms
by Barton, David (DG) 08 Feb '07

08 Feb '07

Feng, Matt, The problem appears to be caused by a error in parsing the Titles portion of this one particular ATOMS.INP file. If you delete the entire reference #2 in the titles area WEBAtoms runs fine. (refer2: Cava, et al, Physica C 165, p.419, (1990)). Error can be reproduced by: 1) Choose YBa2Cu3O7.inp : yttrium barium copper oxide ( 123 , ybco) entry in http://cars9.uchicago.edu/cgi-bin/atoms/Atoms-Search.cgi and run WebAtoms 2) Choose K-edge 3) Run ATOMS (You will see the ERROR ACCESS Denied) The work around: 1) Choose YBa2Cu3O7.inp : yttrium barium copper oxide ( 123 , ybco) entry in http://cars9.uchicago.edu/cgi-bin/atoms/Atoms-Search.cgi and run WebAtoms 2) Choose K-edge 3) Delete (refer2: Cava, et al, Physica C 165, p.419, (1990)) in the header. 4) Run ATOMS (and it now works fine). Dave >-----Original Message----- >From: ifeffit-bounces(a)millenia.cars.aps.anl.gov >[mailto:ifeffit-bounces@millenia.cars.aps.anl.gov] On Behalf >Of Feng Wang >Sent: Wednesday, February 07, 2007 2:01 AM >To: XAFS Analysis using Ifeffit >Subject: Re: [Ifeffit] ERROR in getting feff8 input file for >YBCO from Webatoms > >Thank you, Paul, for clarifying this problem. > >Sincerely, > >Feng > > >> Just to follow on this -- I can access the YBCO atoms entry (via a >> search). The atoms.inp pops up correctly, but when I >attempted to run >> atoms, I got the same error message. The error doesn't come from my >> end (as far as I can tell). The problem seems to be with the >execution >> of atoms (e.g. hitting the button "Run atoms".). The cgi is >probably >> broken. >> >> Paul >> >> ERROR >> >> Access Denied >> >> Access Denied by security policy >> The security policy for your network prevents your request >from being >> allowed at this time. Please contact your administrator if you feel >> this is incorrect. >> >> On Feb 3, 2007, at 2:42 AM, Matt Newville wrote: >> >> > Hi Feng, >> > >> > Sorry for the trouble. I'm not sure where the problem >is. I can get >> > files from the Atoms archive from home (ie, outside the Argonne >> > network), and I don't see any recent errors in the logs, >suggesting >> > that both the server connections and the access through the Argonne >> > and APS firewalls are working. Could it be that the >trouble is at >> > your end? >> > >> > Perhaps we can troubleshoot this a little better. Does the 'search >> > for Atoms.inp' page >> > http://cars9.uchicago.edu/~newville/adb/search.html >> > appear for you? Is it responsive? >> > >> > Can you select a central atoms, a structure, and then hit 'Run >> > Webatoms'? >> > >> > Can you browse to http://cars9.uchicago.edu/atomsdb/ and >see a large >> > number of atoms.inp files? Can you select and download >some of these >> > files? >> > >> > Hope that helps, >> > >> > --Matt >> > _______________________________________________ >> > Ifeffit mailing list >> > Ifeffit(a)millenia.cars.aps.anl.gov >> > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit >> >> _______________________________________________ >> Ifeffit mailing list >> Ifeffit(a)millenia.cars.aps.anl.gov >> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit > > > >==================================== >EM group >Physics & NINT >Phone: (780)6411667 >http://www.phys.ualberta.ca/~fwang >_______________________________________________ >Ifeffit mailing list >Ifeffit(a)millenia.cars.aps.anl.gov >http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit >

4 3

Re: [Ifeffit] ERROR in getting feff8 input file for YBCO from Webatoms
by Feng Wang 07 Feb '07

07 Feb '07

Thank you, Paul, for clarifying this problem. Sincerely, Feng > Just to follow on this -- I can access the YBCO atoms entry (via a > search). The atoms.inp pops up correctly, but when I attempted to > run atoms, I got the same error message. The error doesn't come from > my end (as far as I can tell). The problem seems to be with the > execution of atoms (e.g. hitting the button "Run atoms".). The cgi > is probably broken. > > Paul > > ERROR > > Access Denied > > Access Denied by security policy > The security policy for your network prevents your request from being > allowed at this time. Please contact your administrator if you feel > this is incorrect. > > On Feb 3, 2007, at 2:42 AM, Matt Newville wrote: > > > Hi Feng, > > > > Sorry for the trouble. I'm not sure where the problem is. I can get > > files from the Atoms archive from home (ie, outside the Argonne > > network), and I don't see any recent errors in the logs, suggesting > > that both the server connections and the access through the Argonne > > and APS firewalls are working. Could it be that the trouble is at > > your end? > > > > Perhaps we can troubleshoot this a little better. Does the 'search for > > Atoms.inp' page > > http://cars9.uchicago.edu/~newville/adb/search.html > > appear for you? Is it responsive? > > > > Can you select a central atoms, a structure, and then hit 'Run > > Webatoms'? > > > > Can you browse to http://cars9.uchicago.edu/atomsdb/ and see a large > > number of atoms.inp files? Can you select and download some of these > > files? > > > > Hope that helps, > > > > --Matt > > _______________________________________________ > > Ifeffit mailing list > > Ifeffit(a)millenia.cars.aps.anl.gov > > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit > > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit ==================================== EM group Physics & NINT Phone: (780)6411667 http://www.phys.ualberta.ca/~fwang

1 0

ERROR in getting feff8 input file for YBCO from Webatoms
by Feng Wang 07 Feb '07

07 Feb '07

Hi all, I used Webatoms (in http://cars9.uchicago.edu/~newville/adb/search.html) to get the feff8 input file for YBa2Cu3O7, but was given an error message, as shown below: ERROR Access Denied ------------------------------------------- Access Denied by security policy The security policy for your network prevents your request from being allowed at this time. Please contact your administrator if you feel this is incorrect I tried several other materials, there is no such kind error message. Could you please help with that? Thank you so much in advance. Sincerely, Feng ==================================== EM group Physics & NINT Phone: (780)6411667 http://www.phys.ualberta.ca/~fwang

3 2

Re: [Ifeffit] statistical vicissitudes and dead links
by Kelly, Shelly D. 26 Jan '07

26 Jan '07

Hi Paul, Here is a PDF version of the relevant pages from Dani's thesis. Shelly --------------------------- Shelly Kelly Argonne National Laboratory Bldg 203, RM E113 9700 S. Cass Ave Argonne, IL 60439 Molecular Environmental Science Group www.mesg.anl.gov skelly(a)anl.gov phone: 630-252-7376 > -----Original Message----- > From: ifeffit-bounces(a)millenia.cars.aps.anl.gov [mailto:ifeffit- > bounces(a)millenia.cars.aps.anl.gov] On Behalf Of Paul Fons > Sent: Saturday, January 20, 2007 7:38 PM > To: XAFS Analysis using Ifeffit > Subject: Re: [Ifeffit] statistical vicissitudes and dead links > > Hi Shelly, > I noticed that the link you posted either is dead or doesn't work > from outside anl (the message from the server is ambiguous). > > Paul > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

1 0

Multiple k-weight fitting
by David C. Schmitter 22 Jan '07

22 Jan '07

Hi everyone. I am currently trying to reduce the correlations between N and sigma squared. I have been told that multiple k-weight fitting is the way to go about this; but I am curious what the best method is to try and do this using Artemis. Is it simply a matter of clicking the other weighting options available? If so, what should I be looking for to determine whether my parameters are correct or not? I am not sure if this is a redundant question in terms of the mailing list; but I am a relatively new user and cannot remember seeing anything about it yet. Thanks. Dave Schmitter David C. Schmitter Dept. of Physics and Astronomy Nebraska Center for Materials and Nanoscience University of Nebraska Lincoln dschmitt(a)unlserve.unl.edu (402) 472-5094

2 1

Re: [Ifeffit] statistical vicissitudes
by Kelly, Shelly D. 21 Jan '07

21 Jan '07

Hi Aaron, In EXAFS the expected deviation in RCS(reduced chi square) is sqrt(2/nu) So one fit is better than another fit if their difference is greater than twice the fluctuation in the RCS values (RCS1-RCS2)/RCS2 > 2 sqrt(2/nu) We divide by RCS2 because our RCS-values are poorly scaled. {in your link to the nist pages F = RCS1/RCS2 -1 and the statistically significant variation in F is 2 * sqrt(2/nu)} If there are different values for nu in the two models then (RCS1-RCS2) > 2sqrt(RCS2^2/nu2 + RCS1^2/nu1) I think that the best explanation of these tests are given in Dani Haskel's thesis pg 39 - 41 http://www.aps.anl.gov/xfd/people/haskel/dani.html and Bevington, P. R.; Robinson, D. K. Data Reduction and Error Analysis for the Physical Sciences; 2 ed.; McGraw-Hill: New York, 1992. HTH Shelly > -----Original Message----- > From: ifeffit-bounces(a)millenia.cars.aps.anl.gov [mailto:ifeffit- > bounces(a)millenia.cars.aps.anl.gov] On Behalf Of Aaron Slowey > Sent: Thursday, January 18, 2007 4:50 PM > To: ifeffit(a)millenia.cars.aps.anl.gov > Subject: [Ifeffit] statistical vicissitudes > > Hello - > > The use of R-factors and reduced-chi-squared values, it's computations > by IFEFFIT have been made clear by past writings, but the use of > reduced-chi-squared as a comparative statistic has not, to my > knowledge, been completely well-documented. > > Not withstanding the three disconnects between classical statistics > and the realities of EXAFS mentioned in the IFEFFIT faq section > > http://cars9.uchicago.edu/iffwiki/FAQ/FeffitModeling#head- > ed63cbd00d0b85def31ada3a695873243d08a76a > > can > > Is the criterion provided by Kelly et al. (2001) J. Phys. Chem. B 105(27) > > http://www.mesg.anl.gov/pdf/JPC2001samms.pdf > > namely, redchi-sq_1 / redchi-sq_2) - 1 > 2 * (2 / deg_of_freedom)^0.5, > > derived from the F-test formalism and, if so, how? > > I tried to derive the above equation based on the F-test formalism > provided by NIST: > > http://www.itl.nist.gov/div898/handbook/prc/section3/prc32.htm > > but was unsuccessfull. The Kemner et al. paper referenced by Kelly et > al. to explain the statistical basis for the above equation does not > clearly do so. > > I request that someone please provide a reference to a *first > principles* basis for criteria to compare reduced-chi-square values of > alternate EXAFS fits. > > Aaron > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

2 1