voigt_profile

[yevgenyr]

Hi Dan,
Great code - thanks!

I was wondering if you have plans to add a voigt peak shape (see here).

I remember that realistically, line profiles usually have this lineshape. Gaussians are usually rare.

it can be something like:

t = np.arange(-1000, 1000, 0.1) 
fwhmg = 2  
fwhml = 2 
v = sps.voigt_profile(t, fwhmg, fwhml) 
I = np.convolve(I, v, 'same') 

inside Scatter.generate_powder. (I is the pxrd profile with delta functions)

[DanPorter]

Hi yevgenyr,

Thanks for your message and glad to hear you like the code!

That's a great idea and something I had planned on adding at some point (when someone asked for it). Now you have asked, I'll find a way to add this feature :)

Best wishes,

Dan

[yevgenyr]

Thanks Dan,
If you'de like I can also contribute in a way of Pull Requests as I go.
I am working on integrating a different CIF reader (from diffpy - it is a bit more comprehansive) and map it over to your cif reader. (of course feel free to reject/modify my PRs :))

[DanPorter]

Hi yevgenyr,
I've just committed a new version (V2.2.0) with some new features on the calculation of powder patterns:

twotheta, intensity, reflections = xtl.Scatter.powder('xray', units='twotheta', energy_kev=8, lorentz_fraction=0.5, custom_peak=None)

The are new inputs (all can also be set using xtl.Scatter.setup_scatter):

• units: something like 'Q', 'dspacing' or 'twotheta'
• lorentz_fraction: the ratio of Gaussian/Lorentzian widths in pseudo-Voight
• custom_peak: if an array is given, this is convolved with the intensity delta-functions, rather than the pseudo-Voight.

And there are new outputs:

• twotheta: array of spectrum angles
• intensity: array of spectrum intensity
• reflections: array of grouped reflections (h,k,l,twotheta,intensity)

I've added an example showing this functionality in the examples folder.

Please do have a go and see what you think! If you spot any problems let me know.

Out of interest - what features does the diffpy CIF reader have that aren't in my (admittedly pretty simple) one?