Re: [Ifeffit] EXAFS fitting sum of all paths in R space
...Basically, you need to sum different paths in k-space, save as chi(k), then import to Athena, perform FT in the same range as your data, and then the result should be the same as obtained in Artemis for the fit vs data. That will preserve the phase relationships between different paths, while your procedure does not have that information because you save magnitudes of the paths. Anatoly On Wed, Sep 23, 2020 at 8:58 AM Ashwini Kumar Poswal < poswalashw...@gmail.com> wrote: > Dear All, > I have one problem regarding the sum of all paths (R-space) after fitting. > If we include many paths for the fitting, the fitting shows a good fit but > after fitting, I select all paths and export (using Artemis [Data] window: > save data + marked paths as |χ(R)|. I plotted all paths in origin as well > as the sum of all paths. But when I add up all the contributions from all > the paths, the deviation at larger paths is extraordinarily high and not > equal to final fit. I am attaching a figure for illustration. It is clearly > seen that the final fit is lesser compared to one of the path and the sum > of all paths is quite high compared to fit (as shown in second figure. Can > anybody suggest, how to sum different paths? > [image: image.png] > [image: image.png] > > Regards > *Ashwini Kumar Poswal* > *Mumbai* > *Tel.-+91 (022) 25595081* > ___ > 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 > ___ 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
Re: [Ifeffit] EXAFS fitting sum of all paths in R space
It is a nice demonstration that |χ(R)| is not equal to |χ1(R)|+ |χ2(R)| + ... |χn(R)|. It is equal to: |χ1(R) + χ2(R) + ... χn(R) |. The latter is most often smaller than the former, according to the Cauchy-Schwarz inequality. Anatoly On Wed, Sep 23, 2020 at 8:58 AM Ashwini Kumar Poswal < poswalashw...@gmail.com> wrote: > Dear All, > I have one problem regarding the sum of all paths (R-space) after fitting. > If we include many paths for the fitting, the fitting shows a good fit but > after fitting, I select all paths and export (using Artemis [Data] window: > save data + marked paths as |χ(R)|. I plotted all paths in origin as well > as the sum of all paths. But when I add up all the contributions from all > the paths, the deviation at larger paths is extraordinarily high and not > equal to final fit. I am attaching a figure for illustration. It is clearly > seen that the final fit is lesser compared to one of the path and the sum > of all paths is quite high compared to fit (as shown in second figure. Can > anybody suggest, how to sum different paths? > [image: image.png] > [image: image.png] > > Regards > *Ashwini Kumar Poswal* > *Mumbai* > *Tel.-+91 (022) 25595081* > ___ > 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 > ___ 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
[Ifeffit] EXAFS fitting sum of all paths in R space
Dear All, I have one problem regarding the sum of all paths (R-space) after fitting. If we include many paths for the fitting, the fitting shows a good fit but after fitting, I select all paths and export (using Artemis [Data] window: save data + marked paths as |χ(R)|. I plotted all paths in origin as well as the sum of all paths. But when I add up all the contributions from all the paths, the deviation at larger paths is extraordinarily high and not equal to final fit. I am attaching a figure for illustration. It is clearly seen that the final fit is lesser compared to one of the path and the sum of all paths is quite high compared to fit (as shown in second figure. Can anybody suggest, how to sum different paths? [image: image.png] [image: image.png] Regards *Ashwini Kumar Poswal* *Mumbai* *Tel.-+91 (022) 25595081* ___ 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
