dward
Hi,
I analyze TOF data from a calibration sample. It turns out that I cannot model the peak shape using standard functions, for example a back-to-back convolution. Some peaks in my spectrum have bumps with long tails on the right hand side. It looks like a composition of two peaks. I know that the effect comes from the neutron moderator. Therefore I wanted to ask what is the procedure to define a more complicated convolution? So far the only way to fit this data is adding an additional phase with a slightly different lattice parameter but obviously this is not what I intend.
Best wishes,
Dariusz
johnsoevans
Which diffractometer is the data from? What's the standard? What peak shape are you trying to use?
It would also be worth asking the instrument scientist. They will probably be able to help more than the forum.
dward
Hi John,
I was able to fit the profile function with the macro wifd_mic_new defined, if I am correct, for the GEM instrument. Now I am fitting single peaks (using xo_Is) in order to obtain delta d to d. I am using the same type of the convolution and it works very well. The only thing I need is to export the calculated profile with a higher density of points. When I use the following macro:
xdd_out ######.dat load out_record out_fmt out_eqn
{
" %11.6f " = X;
" %11.6f " = Yobs;
" %11.6f " = Ycalc;
" %11.6f\n" = Yobs-Ycalc;
}
I get only the same number of points as it is in the data file.
johnsoevans
Darius,
Alan may have a better answer to this. In your input file there will be a line containing something like "x_calculation_step" which tells Topas how many points to use for the calculated profile. The default is often to "Get" that information from the data. This is only the calculation though, and won't help you with outputting extra points.
Two ideas:
1. Once you have done fit you could do a zero-iters fit to a dummy dataset which has the x-points you want and output ycalc. Th
2. You could look at a macro like the one below which is used to calculate a pattern. Again, modify your .inp file and run zero iters. Have a look at "simulate_pattern_from_cif" in the jedit menus on how to use this.
John
macro calc_pattern(start,stop,step,filename) {yobs_eqn !aac.xy = 1; min start max stop del step
xdd_out filename load out_record out_fmt out_eqn
{
" %11.5f " = X;
" %11.5f\n" = Ycalc;
}
}
dward
Thanks John,
it works exactly as I wanted!