VXRD: Profiling

[qzhu2017]

@sabaronett

So far, we have four options of profiling functions in the pxrd code.

• Gaussian: fwhm
• Lorentzian: fwhm
• Pseudo-Voigt: fwhm, eta,
• Mod-pseudo-Voigt: fwhm, A, eta_h, eta_l, N https://github.com/qzhu2017/XRD/blob/a02536f2dcf74fd28074cc1bec539212eae1b480/pyxtal_xrd/profile.py#L4-L59

At the moment, we have three options in vxrd:

• Gaussian: fwhm
• Lorentzian: fwhm
• Pseudo-Voigt: This is actually Mod-pseudo-Voigt

Can you try to fix this and make it be consistent with our pard code.

[sabaronett]

OK. I'll try to take care of this today.

[sabaronett]

@qzhu2017, I've updated the profiling options and parameters in 21b8738747c9a2e2bbd3c094aeeef1e1ebef8d00. However, after reviewing pyxtal_xrd/profile.py I want to make sure we're on the same page with the parameters available to the user.

Pseudo-Voigt

Although the function prototype here https://github.com/qzhu2017/XRD/blob/a02536f2dcf74fd28074cc1bec539212eae1b480/pyxtal_xrd/profile.py#L50 shows the argument eta, when you call get_profile for the XRD object https://github.com/qzhu2017/XRD/blob/a02536f2dcf74fd28074cc1bec539212eae1b480/pyxtal_xrd/profile.py#L137-L150 eta is essentially calculated from fwhm and not passed from a kwarg. Thus in 21b8738747c9a2e2bbd3c094aeeef1e1ebef8d00, I display a FWHM field for this profile, but not eta.

Modified pseudo-Voigt

Although the function prototype https://github.com/qzhu2017/XRD/blob/a02536f2dcf74fd28074cc1bec539212eae1b480/pyxtal_xrd/profile.py#L5 shows the argument N, get_profile does this, https://github.com/qzhu2017/XRD/blob/a02536f2dcf74fd28074cc1bec539212eae1b480/pyxtal_xrd/profile.py#L124 and this, https://github.com/qzhu2017/XRD/blob/a02536f2dcf74fd28074cc1bec539212eae1b480/pyxtal_xrd/profile.py#L152-L162 i.e., N and fwhm are calculated from other parameters, and U, V, and W, are accepted as kwargs. Thus 21b8738747c9a2e2bbd3c094aeeef1e1ebef8d00 displays everything as before (when it was the only pseudo-Voigt option).

Let me know if my analysis makes sense, if I'm mistaken, or if you wish to change anything about how we want the profiling stuff to work. Thanks.

[qzhu2017]

@sayred1 Can you double check what are the necessary input parameters for pseudo-Voigt profile function. I think eta is needed. But the current code seems to compute eta from fwhm. Can you show here which formula you used to compute pseudo-Voigt?

[sayred1]

@qzhu2017

So according to https://onlinelibrary.wiley.com/doi/pdf/10.1107/S0021889887087090, eta can be determined to 1% when calculating it in that fashion as a function of the fwhm.

[qzhu2017]

@sabaronett I think you are correct here.