Re: [Ifeffit] How to change the data file in Dartemis?
Dear Bruce, dear all, Just to add some input and precise my question. The problem I encounter with "replace this chi(k)" did not appear to me until I tried to fit a spectrum twice with two different normalisations. In this case, after a fit on the first spectrum, I "replace this chi" with another that is normalised differently, and when I try to fit, it stops with the following error message: This data came from the the same source as another data group. You seem to be trying to increase your number of independent points by fitting the same data more than once in a multiple data set fit. Seeing that, I wondered whether "replace this chi" was doing what I thought. The upgrade did not change anything. I just tried with a colleague working under windows and the result is also the same. Shall I do differently to fit those two really similar spectra? Thank you for your help, Isabelle On 01/07/2016 21:08, Bruce Ravel wrote:
On 07/01/2016 01:57 PM, Isabelle Kieffer wrote:
You mean that when you "Replace this chi(k)", it asks what to do, or it supposes that you want to fit new data with the same model?
What I mean is: upgrade. Many of the problems you are reporting have been fixed already. If you upgrade and still see these problems, make a bug report and I will look into it.
B
-- Isabelle KIEFFER, Ingénieur de Recherche CNRS, Ligne CRG FAME (BM30B) ESRF - The European Synchrotron CS 40220 - 38043 Grenoble Cedex 9 - France Tel : +33 (0)4 76 88 25 47 - Bureau : LOB 2-50
Hi Isabelle, If I understand your emails correctly, you are trying to fit two variations of a dataset (with different normalisations) to a given model FEFF calculation, yes? You have two different chi(k)'s. Rather than trying to replace. have you tried: Just fitting them one at a time in Artemis? i.e. Import the first chi(k), set the FT parameters etc. Generate the FEFF model. Pick paths. Set the GDS values. Save the project file. Run fit. Discard first data, import second set the FT parameters, select paths, enter the GDS values in the data window...(save again, perhaps as different fpj) and fit again? (OK...this is replace, but by brute force) How about fitting both at the same time using the same FEFF model: Import both chi(k)'s and run two independent fits (i.e. amp1 for 1st, amp2 for 2nd etc) with the same FEFF calculation? I have tested both of these routes and they do work (though there is an issue with the Artemis [Plot] window dropping one of the datasets if it isn't being fit). regards, -R. On 7/7/2016 7:04 AM, Isabelle Kieffer wrote:
Dear Bruce, dear all,
Just to add some input and precise my question.
The problem I encounter with "replace this chi(k)" did not appear to me until I tried to fit a spectrum twice with two different normalisations. In this case, after a fit on the first spectrum, I "replace this chi" with another that is normalised differently, and when I try to fit, it stops with the following error message:
This data came from the the same source as another data group. You seem to be trying to increase your number of independent points by fitting the same data more than once in a multiple data set fit.
Seeing that, I wondered whether "replace this chi" was doing what I thought. The upgrade did not change anything. I just tried with a colleague working under windows and the result is also the same.
Shall I do differently to fit those two really similar spectra?
Thank you for your help,
Isabelle
On 01/07/2016 21:08, Bruce Ravel wrote:
On 07/01/2016 01:57 PM, Isabelle Kieffer wrote:
You mean that when you "Replace this chi(k)", it asks what to do, or it supposes that you want to fit new data with the same model?
What I mean is: upgrade. Many of the problems you are reporting have been fixed already. If you upgrade and still see these problems, make a bug report and I will look into it.
B
Robert: Thanks for weighing in. Isabelle: I apologize if it seems like I am ignoring you. I was on travel for the majority of June and am very busy with things not related to software right now. I simply haven't had time to look into your problem. I'll get to it eventually. B On 07/07/2016 07:59 PM, Robert Gordon wrote:
Hi Isabelle,
If I understand your emails correctly, you are trying to fit two variations of a dataset (with different normalisations) to a given model FEFF calculation, yes? You have two different chi(k)'s. Rather than trying to replace. have you tried:
Just fitting them one at a time in Artemis? i.e. Import the first chi(k), set the FT parameters etc. Generate the FEFF model. Pick paths. Set the GDS values. Save the project file. Run fit. Discard first data, import second set the FT parameters, select paths, enter the GDS values in the data window...(save again, perhaps as different fpj) and fit again? (OK...this is replace, but by brute force)
How about fitting both at the same time using the same FEFF model: Import both chi(k)'s and run two independent fits (i.e. amp1 for 1st, amp2 for 2nd etc) with the same FEFF calculation?
I have tested both of these routes and they do work (though there is an issue with the Artemis [Plot] window dropping one of the datasets if it isn't being fit).
regards, -R.
On 7/7/2016 7:04 AM, Isabelle Kieffer wrote:
Dear Bruce, dear all,
Just to add some input and precise my question.
The problem I encounter with "replace this chi(k)" did not appear to me until I tried to fit a spectrum twice with two different normalisations. In this case, after a fit on the first spectrum, I "replace this chi" with another that is normalised differently, and when I try to fit, it stops with the following error message:
This data came from the the same source as another data group. You seem to be trying to increase your number of independent points by fitting the same data more than once in a multiple data set fit.
Seeing that, I wondered whether "replace this chi" was doing what I thought. The upgrade did not change anything. I just tried with a colleague working under windows and the result is also the same.
Shall I do differently to fit those two really similar spectra?
Thank you for your help,
Isabelle
On 01/07/2016 21:08, Bruce Ravel wrote:
On 07/01/2016 01:57 PM, Isabelle Kieffer wrote:
You mean that when you "Replace this chi(k)", it asks what to do, or it supposes that you want to fit new data with the same model?
What I mean is: upgrade. Many of the problems you are reporting have been fixed already. If you upgrade and still see these problems, make a bug report and I will look into it.
B
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
-- Bruce Ravel ------------------------------------ bravel@bnl.gov National Institute of Standards and Technology Synchrotron Science Group at NSLS-II Building 743, Room 114 Upton NY, 11973 Homepage: http://bruceravel.github.io/home/ Software: https://github.com/bruceravel Demeter: http://bruceravel.github.io/demeter/
No problem... I was just not sure of what I was observing and feared I was doing wrong from the begining. The solution of Robert is working and if I am using correctly the "Replace this chi(k)" in the normal cases (which means that it does not take into account the previous dataset when it starts fitting the second), I will continue using it. For different chi's coming from the same data, I will use the so called "brute force" method for the moment. Thank you to both of you. Isabelle On 08/07/2016 14:54, Bruce Ravel wrote:
Robert: Thanks for weighing in.
Isabelle: I apologize if it seems like I am ignoring you. I was on travel for the majority of June and am very busy with things not related to software right now. I simply haven't had time to look into your problem. I'll get to it eventually.
B
On 07/07/2016 07:59 PM, Robert Gordon wrote:
Hi Isabelle,
If I understand your emails correctly, you are trying to fit two variations of a dataset (with different normalisations) to a given model FEFF calculation, yes? You have two different chi(k)'s. Rather than trying to replace. have you tried:
Just fitting them one at a time in Artemis? i.e. Import the first chi(k), set the FT parameters etc. Generate the FEFF model. Pick paths. Set the GDS values. Save the project file. Run fit. Discard first data, import second set the FT parameters, select paths, enter the GDS values in the data window...(save again, perhaps as different fpj) and fit again? (OK...this is replace, but by brute force)
How about fitting both at the same time using the same FEFF model: Import both chi(k)'s and run two independent fits (i.e. amp1 for 1st, amp2 for 2nd etc) with the same FEFF calculation?
I have tested both of these routes and they do work (though there is an issue with the Artemis [Plot] window dropping one of the datasets if it isn't being fit).
regards, -R.
On 7/7/2016 7:04 AM, Isabelle Kieffer wrote:
Dear Bruce, dear all,
Just to add some input and precise my question.
The problem I encounter with "replace this chi(k)" did not appear to me until I tried to fit a spectrum twice with two different normalisations. In this case, after a fit on the first spectrum, I "replace this chi" with another that is normalised differently, and when I try to fit, it stops with the following error message:
This data came from the the same source as another data group. You seem to be trying to increase your number of independent points by fitting the same data more than once in a multiple data set fit.
Seeing that, I wondered whether "replace this chi" was doing what I thought. The upgrade did not change anything. I just tried with a colleague working under windows and the result is also the same.
Shall I do differently to fit those two really similar spectra?
Thank you for your help,
Isabelle
On 01/07/2016 21:08, Bruce Ravel wrote:
On 07/01/2016 01:57 PM, Isabelle Kieffer wrote:
You mean that when you "Replace this chi(k)", it asks what to do, or it supposes that you want to fit new data with the same model?
What I mean is: upgrade. Many of the problems you are reporting have been fixed already. If you upgrade and still see these problems, make a bug report and I will look into it.
B
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
-- Isabelle KIEFFER, Ingénieur de Recherche CNRS, Ligne CRG FAME (BM30B) ESRF - The European Synchrotron CS 40220 - 38043 Grenoble Cedex 9 - France Tel : +33 (0)4 76 88 25 47 - Bureau : LOB 2-50
participants (3)
-
Bruce Ravel -
Isabelle Kieffer -
Robert Gordon