alan-jarvis
Hi is me again,
I am using the macro Create_hklm_d_Th2_Ip_file to give me some output.
I'd really like to get more output on my peaks.
Is there another macro that I can use to get for each refined peak:
- Phase name (where it is generated using hkl_Is lebail)
- fwhm
- error in peak fit
- peak area
- Lorentzian/Gaussian fraction (i.e. pv_lor)
- number of data points per peak
In fact, is it possible to get an output of all of the refined variable/parameters?
That would be very useful if you're doing a lot of refinements.
EXPGUI/GSAS gives this sort of output, I would think that this sort of thing is also useful and possible from Topas since both are Rietveld refinements?
Maybe I can write a macro for it: has anyone ever written any tutorial or notes on how to write a Jedit/Topas macro?
Maybe this is an option in Topas_Brukker, I dunno.
I'm from the wrong side of the tracks, so I don't know the rich folks who own that software... :blush:
Thanks,
Alan Jarvis
Sheffield
alancoelho
The INP file is the reference for what parameters is refined or not. If you use jEdit then refined parameters are shown in Red.
alan-jarvis
Thank you for the answer but that was not my question.
I understand that when refined the results are in RED in Jedit.
But they are stuck in Jedit.
It seems pretty cumbersome to manually copy and paste the results from refining from the Jedit file.
Can one get output of these in some format like a text file?
As I said, I am using Create_hklm_d_Th2_Ip_file to give (limited) output.
I want:
- Phase name (where it is generated using hkl_Is lebail)
- fwhm
- error in peak fit
- peak area
- Lorentzian/Gaussian fraction (i.e. pv_lor)
And ideally:
- number of data points used for the refining per peak
(I would think that the program does a check on this? to make sure you get valid results.
After all, there's a lower limit to the number of data points to get a valid refinement per peak).
I would like output of ALL of the refined variable/parameters.
alancoelho
>It seems pretty cumbersome to manually copy and paste the
>results from refining from the Jedit file.
>Can one get output of these in some format like a text file?
The INP file is a text file. Anyway, you can probably output what you want using the 'out' keyword but there's no automatic means of outputting values.
Some of the information you seek are difficult to determine and IMO not necerssary.
- fwhm: This does not make much sense in the context of an emission profile convoluted by a Voigt for example. The Voigt would have a FWHM but the actual peak would comprise a number of emission profile lines plus other convolutions. Thus the idea of a FWHM is not meaningful.
- error in peak fit: I’m not sure what you mean by this. Use ‘do_errors’ to see the correlation matrix
- peak area: explained in another post
- Lorentzian/Gaussian fraction (i.e. pv_lor): this is pv_lor
alancoelho
>And ideally:
>- number of data points used for the refining per peak
>(I would think that the program does a check on this? to make sure you get valid results.
>After all, there's a lower limit to the number of data points to get a valid refinement per peak).
Thought I should elaborate. Yes, I could see this being a little useful if you were fitting pseudo-Voigts or say PVII peaks to non-XRD type data. In practice peaks are fitted using an instrument function as well as an emission profile plus a specimen contribution. They ensure that peaks have non-trivial broadening. Typically also a fit should be visually analysed in detail with the separate plot for individual peaks displayed. On top of all that the errors and correlations in peak parameters should be inspected. I can also only add that I have never had a need for such information.
alancoelho
I forgot; all independent parmaeter values can be outputted using:
out_prm_vals_per_iteration
or,
out_prm_vals_on_convergence
out_prm_vals_filter can be used to output filtered parameter values.
For example,
out_prm_vals_on_convergence file.out append
out_prm_vals_filter "a* b*"
This outputs all parameters starting with the letter 'a' or 'b'.
Version 5 can also output dependent parameters using out_prm_vals_dependents_filter. The test_examples\out_prm_vals.inp example has:
out_prm_vals_per_iteration out_prm_vals-iter.txt
out_prm_vals_filter "stl csl b* !bkg*"
out_prm_vals_dependents_filter "dependent* !dependent3"
out_prm_vals_on_convergence out_prm_vals-conv.txt
out_prm_vals_filter "stl csl b* !bkg*"
out_prm_vals_dependents_filter "dependent* !dependent3"
XDD(ceo2)
CuKa5(0.0001)
Radius(173)
LP_Factor(@, 3)
Full_Axial_Model(12, 20, 12, 5.1, @ 7)
Divergence(1)
Slit_Width(.1)
bkg @ 0 0 0 0 0
One_on_X(@, 1000)
Zero_Error(@, 0)
str
space_group F_M_3_M
scale @ 0.0014503208
Cubic(@ 5.41)
site Ce1 occ Ce+4 1 beq b1 0.2028
site O1 x 0.25 y 0.25 z 0.25 occ O-2 1 beq b2 0.5959
CS_L(csl, 300)
Strain_L(stl, 0.1)
Strain_G(@, 0.1)
prm dependent1 = b1 + b2;
prm dependent2 = b1 - b2;
prm dependent3 = b1 * b2;
The file out_prm_vals-conv.txt will then contain:
Cycle Iter Rwp b1 b2 csl stl dependent1 dependent2
0 15 6.926878e+000 2.275081e-001 4.484624e-001 2.590170e+002 1.666311e-002 6.759704e-001 -2.209543e-001
Note, that Cycle, Iter, and Rwp are also outputted.