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

Re: [Ifeffit] What confidence limit is implied in the error bars forleast-squares fitting in Athena
by Don R Baker 03 Oct '08

03 Oct '08

Hi Shelly, I am very interested in Andrew's question too, but I am to ignorant to fully understand your response. Could you please provide some more detail? For example, when you state "a large source of uncertainty can come from the representation of the standards in matching the species in the sample." do you mean that heterogeneities (as Bruce mentioned in his email) in the standards are propagated into the fits of the unknowns? Or instead do you mean that an inappropriate choice of standards can create unreliability in the fit? I would appreciate any details you can add to your repsonse. Thanks, Don On Thu, 2008-02-10 at 14:22 -0700, Kelly, Shelly wrote: > Hi Andrew, > > > > If you are fitting XANES spectra with standards, a large source of > uncertainty can come from the representation of the standards in > matching the species in the sample. > > > > Shelly > > > > > ______________________________________________________________________ > > From: ifeffit-bounces(a)millenia.cars.aps.anl.gov > [mailto:ifeffit-bounces@millenia.cars.aps.anl.gov] On Behalf Of Andrew > Sent: Thursday, October 02, 2008 3:49 PM > To: ifeffit(a)millenia.cars.aps.anl.gov > Subject: [Ifeffit] What confidence limit is implied in the error bars > forleast-squares fitting in Athena > > > > Hi everyone, > > > > I’m fitting data with the least-squares method in Athena and am not > sure what the error bars correspond to. What I mean, are these 95% > confidence limits, or is it some other statistical way for error > analysis? If Athena says the weight of one phase is 0.049 (0.003), > what confidence are in the 0.003 error bar reported by Athena? > > > > (I have done a PCA of the XANES, and XRDs of the samples and am pretty > confident in the phases that I’m fitting). > > > > I greatly appreciate the help! > > > > Andrew Campos > > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit -- Melting rocks today for a better tomorrow . . . Don R. Baker, Professor of Geochemistry, Earth and Planetary Sciences, McGill University, Montreal, QC CANADA phone: 1-514-398-7485

2 1

What confidence limit is implied in the error bars for least-squares fitting in Athena
by Andrew 02 Oct '08

02 Oct '08

Hi everyone, I'm fitting data with the least-squares method in Athena and am not sure what the error bars correspond to. What I mean, are these 95% confidence limits, or is it some other statistical way for error analysis? If Athena says the weight of one phase is 0.049 (0.003), what confidence are in the 0.003 error bar reported by Athena? (I have done a PCA of the XANES, and XRDs of the samples and am pretty confident in the phases that I'm fitting). I greatly appreciate the help! Andrew Campos

5 7

Re: [Ifeffit] Ifeffit Digest, Vol 68, Issue 3
by Frommer Jakob 02 Oct '08

02 Oct '08

Hi Renske, I still think your problem is related to the potentials. If you unzip your artemis project and look in one of the feffXXXX.dat files in the folder data0.feff3 (this calculations was done with the solvent_O) you will find the calculated R_MTs in the first lines - for ipot3 (O) this is 1.6. Do the same with the data0.feff7 folder (apparently this calculation was done without the solvent O - even though the feff.inp says something different) and you will find 0.728. FEFF (always?) uses the closest atom (relative to the ipot=0-atom) to determine the unique potentials to be used for all atoms with the same ipot (in the manual it is stated that FEFF takes the first atom in the atoms-list but in your case just changing the position in the list did change nothing whereas increasing the distance to 3.3 (the O_t_1 atoms are at 3.266) did yield a different R_MT - but maybe I am wrong). HTH Jakob ___________________________________________________ Jakob Frommer Soil Chemistry Group Institute of Biogeochemistry and Pollutant Dynamics, ETH Zürich Universitätstrasse 16, CHN F19 CH - 8092 Zürich Tel: +41 44 632 87 58 Fax: +41 44 633 11 18 web: http://www.ibp.ethz.ch/research/soilchemistry ________________________________ From: ifeffit-bounces(a)millenia.cars.aps.anl.gov on behalf of ifeffit-request(a)millenia.cars.aps.anl.gov Sent: Thu 10/2/2008 06:37 To: ifeffit(a)millenia.cars.aps.anl.gov Subject: Ifeffit Digest, Vol 68, Issue 3 Send Ifeffit mailing list submissions to ifeffit(a)millenia.cars.aps.anl.gov To subscribe or unsubscribe via the World Wide Web, visit http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit or, via email, send a message with subject or body 'help' to ifeffit-request(a)millenia.cars.aps.anl.gov You can reach the person managing the list at ifeffit-owner(a)millenia.cars.aps.anl.gov When replying, please edit your Subject line so it is more specific than "Re: Contents of Ifeffit digest..." Today's Topics: 1. Re: Athena Peak Fitting (Matt Newville) 2. Re: Independence of scattering paths? (van der Veen Renske) ---------------------------------------------------------------------- Message: 1 Date: Wed, 1 Oct 2008 16:23:34 -0500 From: "Matt Newville" <newville(a)cars.uchicago.edu> Subject: Re: [Ifeffit] Athena Peak Fitting To: "XAFS Analysis using Ifeffit" <ifeffit(a)millenia.cars.aps.anl.gov> Message-ID: <b8522e3d0810011423g68bb1b81g143ed34a0569a471(a)mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 Hi Jenny, I would say that method 2 looks better, as the fit over the whole of the specified fitting range is clearly better. I wouldn't necessarily say that "peak 1" in this fit -- I'm assuming you mean the highest energy peak at ~2488.3eV -- is too broad. At that high energy (definitely into the EXAFS region), the idea of peaks coming from specific electronic transitions breaks down, and EXAFS oscillations can be that broad in energy. It seems that you need this "broad peak" because you're fitting out past 2500 eV. I think that's OK (that third broad peak may not be a single electronic transition but is still useful for identifying species). But, if you'd rather not have to use the broader peak, you might also try fitting just the two sharper peaks at lower energy, say out to 2485 eV. Hope that helps, --Matt On Wed, Oct 1, 2008 at 1:31 PM, Jenny Cai <jennyh_cai(a)yahoo.ca> wrote: > Hi Everyone, > > I am using Athena peak fitting function to fit the sulfur spectrum of my > coke sample. I checked everywhere, but couldn't get detailed instruction for > using this function. So, I attached a file with this email to get help. > Could you please take a look at it, and tell me if I did the fitting right > (in terms of fitting procedure only)? > > I did the fitting in two ways, and I know both of them have problems. For > instance, in the first case, I can't get a good step line and have to cut > off the tail of the spectrum. In the second one, peak 1 is way too broad, > but if I cut it off, the fitting will be very poor. So, please tell me which > way is better, or you have some ideas to improve the fitting. > > Thank you in advance! > > > Best, > > Jenny > > ________________________________ > > Yahoo! Canada Toolbar : Search from anywhere on the web and bookmark your > favourite sites. Download it now! > > _______________________________________________ > Ifeffit mailing list > Ifeffit(a)millenia.cars.aps.anl.gov > http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit > > ------------------------------ Message: 2 Date: Thu, 2 Oct 2008 06:36:48 +0200 From: "van der Veen Renske" <Renske.vanderveen(a)psi.ch> Subject: Re: [Ifeffit] Independence of scattering paths? To: <ifeffit(a)millenia.cars.aps.anl.gov> Message-ID: <67F374811ED3544C8543E6C0EE5EAA4B02AFEC8E(a)MAILBOX0B.psi.ch> Content-Type: text/plain; charset="us-ascii" Dear Bruce and Jakob Thanks for your replies. I realize that I should have been a bit more precise in where I actually put this extra atom: it is placed outside the molecule at an axial position of the absorbing Pt atom, representing an ethanol solvent molecule that coordinates. It is not placed between any other atoms that are included in the fit/sum, so the MT potentials should not be disturbed. Thanks to the nice presentation of scattering paths in Artemis, I am sure that I chose the same scattering paths involving the same atoms for both sums. I attached an Artemis project to make things more clear. Thanks a lot & best wishes, Renske > > On Friday 26 September 2008 22:38:58 van der Veen Renske wrote: > > I have a question concerning the independence of scattering > pathways. > > I fitted the EXAFS signal of a molecule consisting of 30 atoms (the > > related FEFF calculation is called calc 1). To see the > effect of the > > surrounding solvent molecules in the chi(k) data, I added > an O-atom in > > the coordinate list of the FEFF input file at an arbitrary distance > > from the absorbing atom (in this case Pt) and I re-ran FEFF > (calc 2). > > As a check within calc 2, I generated the sum of the same > scattering > > paths that were used in the fitting model of calc 1 using > the therein > > obtained structural fit parameters. The new O-atom was not > included in this sum. > > I expected the sum within calc 2 to be identical to the fit > of calc 1 > > based on the fact that the original atoms can not sense the > presence > > of the new O-atom (independence of scattering paths). BUT: the > > resulting sum was significantly different from the fit! > > It is a little hard to answer this question since you > described it to us rather than showing it to us with project > file(s). I have a couple of suggestions, but without being > able to examine what you did, I cannot be sure that either is > relevant. (You can insert here my standard diatribe on the > topic of how to ask questions.) > > One suggestion is to be very careful that you are including > the same set of paths in the calculation. When you add an > atom to a feff.inp file, the indexing of the paths change in > a way that may be hard to predict. Just blindly summing up > the first N paths may not do what you want. > > The other suggestion is to check that you did not place the > oxygen atom in a position that dramatically altered the > muffin tin radii. > The potantials calculation can change quite if you insert an > atom at an arbitrary loctation such that the muffin tin radii > have to be very different for some atoms. > > B > > > -- > Bruce Ravel ------------------------------------ bravel(a)bnl.gov > > National Institute of Standards and Technology Synchrotron > Methods Group at NSLS --- Beamlines U7A, X24A, X23A2 > Building 535A Upton NY, 11973 > > My homepage: http://xafs.org/BruceRavel > EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/ >

1 0

Re: [Ifeffit] Independence of scattering paths?
by van der Veen Renske 01 Oct '08

01 Oct '08

Dear Bruce and Jakob Thanks for your replies. I realize that I should have been a bit more precise in where I actually put this extra atom: it is placed outside the molecule at an axial position of the absorbing Pt atom, representing an ethanol solvent molecule that coordinates. It is not placed between any other atoms that are included in the fit/sum, so the MT potentials should not be disturbed. Thanks to the nice presentation of scattering paths in Artemis, I am sure that I chose the same scattering paths involving the same atoms for both sums. I attached an Artemis project to make things more clear. Thanks a lot & best wishes, Renske > > On Friday 26 September 2008 22:38:58 van der Veen Renske wrote: > > I have a question concerning the independence of scattering > pathways. > > I fitted the EXAFS signal of a molecule consisting of 30 atoms (the > > related FEFF calculation is called calc 1). To see the > effect of the > > surrounding solvent molecules in the chi(k) data, I added > an O-atom in > > the coordinate list of the FEFF input file at an arbitrary distance > > from the absorbing atom (in this case Pt) and I re-ran FEFF > (calc 2). > > As a check within calc 2, I generated the sum of the same > scattering > > paths that were used in the fitting model of calc 1 using > the therein > > obtained structural fit parameters. The new O-atom was not > included in this sum. > > I expected the sum within calc 2 to be identical to the fit > of calc 1 > > based on the fact that the original atoms can not sense the > presence > > of the new O-atom (independence of scattering paths). BUT: the > > resulting sum was significantly different from the fit! > > It is a little hard to answer this question since you > described it to us rather than showing it to us with project > file(s). I have a couple of suggestions, but without being > able to examine what you did, I cannot be sure that either is > relevant. (You can insert here my standard diatribe on the > topic of how to ask questions.) > > One suggestion is to be very careful that you are including > the same set of paths in the calculation. When you add an > atom to a feff.inp file, the indexing of the paths change in > a way that may be hard to predict. Just blindly summing up > the first N paths may not do what you want. > > The other suggestion is to check that you did not place the > oxygen atom in a position that dramatically altered the > muffin tin radii. > The potantials calculation can change quite if you insert an > atom at an arbitrary loctation such that the muffin tin radii > have to be very different for some atoms. > > B > > > -- > Bruce Ravel ------------------------------------ bravel(a)bnl.gov > > National Institute of Standards and Technology Synchrotron > Methods Group at NSLS --- Beamlines U7A, X24A, X23A2 > Building 535A Upton NY, 11973 > > My homepage: http://xafs.org/BruceRavel > EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/ >

1 0

update for Windows, fixing problems with Athena/Artemis
by Matt Newville 01 Oct '08

01 Oct '08

Hi Folks, I just put up a Windows updater with new versions of Athena and Artemis which (should / seems to) fix the "Athena is stuck" problems seen be various people. Details on what else changed are below. All windows users with Ifeffit 1.2.10 and 1.2.11 should install this update. If you're still using an older version, beware that this update will be binary incompatible with that version, and you should upgrade to 1.2.11 before installing this update. If you've been having troubles with Athena/Artemis "getting stuck" and these persist after installing this update, please let us know!! --Matt =========== This update to Ifeffit 1.2.11 for Windows corrects several problems with the prior release related to building the Athena et al executables against the wrong version of the Ifeffit library. This update includes newly rebuilt executables for Athena, Artemis, Hephaestus, and Atoms. Additionally, the following features have been added to Athena: 1. Filename are sorted when importing multiple files. 2. Corrected a problem related to importing chi(k) data that do not start at 0. 3. Filetype plugins have been added for: - SSRL binary files - SSRL ASCII files - SSRL MicroEXAFS files - files from the CLS HXMA beamline - files from Photon Factory beamline 12C Enjoy! Bruce Ravel 25 Sep. 2008 ===========

3 3

Broadening
by Eugenio Otal 01 Oct '08

01 Oct '08

Hello everybody, I have a doubt about the LDOS broadening, in xmu.dat I found that core-hole broadening is 4.163, which value should I use in LDOS broadening to reproduce the experimental spectra. Thanks, euG

2 1

Independence of scattering paths?
by van der Veen Renske 01 Oct '08

01 Oct '08

Hello! I have a question concerning the independence of scattering pathways. I fitted the EXAFS signal of a molecule consisting of 30 atoms (the related FEFF calculation is called calc 1). To see the effect of the surrounding solvent molecules in the chi(k) data, I added an O-atom in the coordinate list of the FEFF input file at an arbitrary distance from the absorbing atom (in this case Pt) and I re-ran FEFF (calc 2). As a check within calc 2, I generated the sum of the same scattering paths that were used in the fitting model of calc 1 using the therein obtained structural fit parameters. The new O-atom was not included in this sum. I expected the sum within calc 2 to be identical to the fit of calc 1 based on the fact that the original atoms can not sense the presence of the new O-atom (independence of scattering paths). BUT: the resulting sum was significantly different from the fit! Am I overlooking something here? Aren't the calculated backscatter amplitudes and phases independent from each other? I am using Artemis version 0.8.007, Windows XP, Ifeffit 1.2.9. Thank you! Best regards, Renske Renske M. van der Veen PhD student EPFL / PSI Paul Scherrer Institut Swiss Light Source WSLA/210 5232 Villigen PSI Switzerland Phone: +41 56 310 5362 Fax: +41 56 310 3151 E-mail: renske.vanderveen(a)psi.ch Web: http://lsu.epfl.ch http://femto.web.psi.ch

2 1

Re: [Ifeffit] Independence of scattering paths? (van der Veen Renske)
by Frommer Jakob 30 Sep '08

30 Sep '08

Hi Renske, I am not sure whether I fully understand your question but here is my *two-cent-explanation*: FEFF's first step is to calculate the potentials - since it is a real space code based on the muffin-tin approximation the surrounding of an atom determines the actual scattering potential (even if you are not doing a self-consistent calculation). Since the potentials in turn determine the backscattering amplitudes and phase shifts, placing an additional atom randomly in a structure is certainly not a good idea. A possible work-around may be to find a *similar* (but known) structure with oxygen ligands and combine these two calculations in artemis (for multiple scattering however the geometry can matter a lot). If your solvent contains oxygen already you may (at least as a first approximation and for single scattering) just duplicate that path in artemis. Regards Jakob ___________________________________________________ Jakob Frommer Soil Chemistry Group Institute of Biogeochemistry and Pollutant Dynamics, ETH Zürich Universitätstrasse 16, CHN F19 CH - 8092 Zürich Tel: +41 44 632 87 58 Fax: +41 44 633 11 18 web: http://www.ibp.ethz.ch/research/soilchemistry

1 0

plot linetypes option in athena windows version
by Malte Schulz 22 Sep '08

22 Sep '08

Hi there, I would like to export graphics from Athena to use them in my b&w printed thesis. Therefore it would be really useful to have Athena cycle through linetypes instead of colors. I found the "linetypes" option in the "plot" section of the preferences menu. However it doesn't seem to have any effect. I am using Athena v. 0.8.056 for windows. Malte

4 5

export from large Athena projects
by shiryaev＠crys.ras.ru 19 Sep '08

19 Sep '08

Hi, It seems that there is a problem in exporting normalised data files from large (i.e. >10 curves) Athena projects: Exported data are cut in energy. When one exports the files one by one, this problem does not occurs. All the best, Andrei

6 8