fixing end points of the background fitting

matsmcfly1

something that I can struggle with in Topas at times is fitting a nice background - does anyone have any tips/tricks to help with obtaining a nice background fit? Often I will have a decent fit, but then at the end (or start) it will be way to high (or low); in this situation is it possible to make the end points = the data in order to constrain the fit a little?

P.S. any tips/tricks would be greatly received 🙂

rowlesmr

Which version of TOPAS are you using? there is the keyword/macro "bkg_at" which can be used to constrain bkg values.

johnsoevans

There are also a few tips on the wiki. e.g. if you want to draw your background try: https://topas.awh.durham.ac.uk/doku.php?id=background_straight_line_segments
There are various user-contributed background-related macros at:
https://topas.awh.durham.ac.uk/doku.php?id=macros

The poor background at low/high angles could just be using too many terms in the background polynomial. If you find you need lots of terms to describe the background with the standard Chebychev polynomial you could try fitting a few broad peaks (Gaussian or Lorentzian) plus a low order Chebychev. You can often fit complex backgrounds with a relatively small number of parameters this way. I typically put a minimum fwhm of e.g. 1 degree on the peak to stop it fitting Bragg reflections.