[SOLVED] Background output

ahhill

Dear all.

I would like to output the background along with xdd_out. In the TOPAS gui there is an option to view the background, however this doesn't include Background functions using fit_objects, e.g. PV(a, xo, fwhm, g).

Has anyone implemented this output? Would it be adaptable to output the profiles of individual phases too? i.e. it would be possible to see all the individual components that make up y_calc

Once again, thanks for all responses, and I hope that this will be my last TOPAS forum question this week!

Adrian

johnsoevans

Alan can give the formal answer, but this email appeared on the Rietveld list this week:

The way to let TOPAS display background including "fit_objs" is adding a dummy phase (the phase dose NOT in your sample and fix the "scale" at 0).
Whenever you want to have a look at your background including "fit_objs", just fix the chebychev coefficients and 1/x coefficients, bracket all your real phase with /* */, and press F6.

Yours Sincerely
-------------------
Xiaodong (Tony) Wang
Department of Imaging and Applied Physics Curtin University

ahhill

sorry John, I should have said I tried this method, and yes it does work.

Just wondering if there is a more elegant way.

alancoelho

In GUI mode there’s a “Use Phase” switch. You can turn these ON/OFF as you please using the “Structures/hkl Phases” tab. To then obtain what’s left then Perform a Single Step Refinement, then press F4 and Capture the calculated pattern; this creates a new range which you can fit against or save.

In Launch mode I would simply comment out what I don’t want. If it was something being done a lot then I would do something like:

#define BKG_OUT_ ' comment this line out for normal refinement
#ifdef BKG_OUT_ iters 0 #endif
xdd...
bkg...
One_on_X...
fit_obj...
#ifndef BKG_OUT_
str...
str...
hkl_Is...
xo_Is....
#endif

I could include a new phase dependent keyword such as “for_bkg” which would signal that a phase is for background. I would then need to have a new reserved parameter name such as BKG. The user would then need to include these in the relevant places in order to do the required output. In the end I myself wouldn’t simply use /* */ on what I don’t want. Typically too many options that do not increase functionality makes a program more difficult to use.

I am always open to suggetions but the case for a new option needs to be convincing.

ahhill

Thanks Alan

I con confirm this works fine for a simple showing of the background.

I altered it slightly, just to reduce the number of input lines:

iters 100000
chi2_convergence_criteria 0.001
do_errors

'#define plot_background_ ' comment this line out for normal refinement

xdd...
   bkg...
   fit_obj...

#ifdef plot_background_
      iters 0

      #else
      str...
      str...
#endif

e.g.

I am working on a simplified way of outputting the background fit to an ascii file and will post the resulting macros onto the macro page of the Topas wiki.

ahhill

I have made a guide to outputting all details for publication quality plots using EXPGUI.

More information can be found on this link.

Thanks

Adrian

thomas1

ahhill:1322568515 wrote I have made a guide to outputting all details for publication quality plots using EXPGUI.
More information can be found on this link.

Hello,

Is the script linked supposed to work with the Academic version? Refining with the activated output leads to a abnormal termination stating that TC.exe does not know xdd_out.
I would like to have this data for processing with Gnuplot.

thank you,

Thomas