mfisch
I would like to refine Ruby fluorescence spectra in Topas (this is basically a peak doublet which can be nicely fitted with two pseudo voigt peaks). From the difference in peak position of a sample and the position of a reference ruby, I am currently calculating the pressure directly within Topas. One parameter that influences the final result is temperature. Is it somehow possible to give a fixed esd to temperature (e.g 24 °C +/- 1K) which is then propagated to the final result? Can this be done by penalties or restraints? The temperature is not a refined parameter...
johnsoevans
Sorry, no answer from me. The only thing I could think would be equations to propagate errors manually but that would need a "get error" type command which I don't know....perhaps only Alan would. You could, of course, do that post refinement in e.g. excel.
rowlesmr
The nice thing about doing it in topas is that it would take care of all the parameter correlations and the like.
mfisch
@ John: To get an error is Get_Prm_Error(parametername). Even if i propagate the error manually, I would need a function like Put_Prm_Error... The problem is putting the error back to the final result - or, of course, I can extract it as a separate parameter.
@ Matthew: Exactly... But there are no "real" correlation (least squares), as the parameter is not refined.
alancoelho
Hi Martin
I think I know what you want, you want to define a parameter that is not refined; ie.
prm !t 0.123_0.333
where you have entered the error as 0.333. The parameter is fixed but you would like its error propagated through out the esd calculations.
I think I can do that; could you contact me by e-mail.
cheers
alan