Dear iffefit user,
I found on a paper (E.A. Stern et al. /Physica B 208&209 (1995) 117 120) the definition of R quality factor as:
R_factor ≡ < ∆chi(R space))> = √[ ∑ |chi_C(Ri) – chi_E(Ri)|2 / ∑( chi_E(Ri))2] ----> formula 3
Where Chi(Ri) is a complex function (imaginary and real part of the XAFS spectra in the R space), C concerns the calculated XAFS spectrum, while E refers to the experimental XAFS spectrum.
In the case of a linear combination fitting on the k space performed by ATHENA, the <∆chi> needs a correction. Is it correct to write:
R_factor ≡ < ∆chi(k space))> = √[ ∑ (chi_C(ki) – chi_E(ki))2 / ∑( chi_E(ki))2] ?
Where Chi_C(ki) = x1*chi1(k)+x2*chi2(k) : chi1 and chi2 are the EXAFS functions of the two reference samples used for the linear combination procedure and x1, x2 (with 0
On Fri, Apr 17, 2009 at 8:08 AM, Cammelli Sebastiano
In the case of a linear combination fitting on the k space performed by ATHENA, the <∆chi> needs a correction. Is it correct to write:
R_factor ≡ < ∆chi(k space))> = √[ ∑ (chi_C(ki) – chi_E(ki))2 / ∑( chi_E(ki))2] ?
Where Chi_C(ki) = x1*chi1(k)+x2*chi2(k) : chi1 and chi2 are the EXAFS functions of the two reference samples used for the linear combination procedure and x1, x2 (with 0
Yes, I that is correct. For fits in k-space, the chi data is often k-weighted, in which case all the 'chi' functions in these formulae should be the k-weighted chi(k). --Matt
participants (2)
-
Cammelli Sebastiano
-
Matt Newville