mfisch
Hi all,
In the Tech. Ref. it is written that one should use rebin_with_dx_of instead of x_calculation_step for constant step .xye data (Page 41, V5).
When I refine data with a constant stepsize of 0.0036 deg/step together with the rebin command, Topas gives this message:
Constant x-axis step size detected for xy/xye file
Internally changing to constant x-axis step size for efficiency
(-> this seems somehow contradicting because it is already constant?)
and:
Rebinning yobs to constant step data with a step size of 0.0036
Number of data points before rebinning 33557
Number of data points after rebinning 33554
My data starts at 0.5868 and ends at 121.3884 deg2Th, which is in agreement with the observed number of steps before rebinning.
When outputting the same data with
xdd_out file.txt load out_record out_fmt out_eqn
{
" %9.0f" = Xi;
" %11.5f" = X;
" %11.5f" = Yobs;
}
X starts at 0.5886 and ends at 121.3794, which is in agreement with the observed number of steps after rebinning.
So not only yobs but also X gets interpolated even though there is no need to? Why is it still better to use rebin instead of calculation_step (with which the mentioned interpolation does not happen)?
Thanks,
Martin
alancoelho
Hi Martin
>(-> this seems somehow contradicting because it is already constant?)
XY/XYE data files comprise many x values and y values (ie. XY data). In Version 4.1 the program treated XY files as if the x-axis was in unequal steps; this has some speed drawbacks. In Version 5 the x-axis is checked on loading XY/XYE files and if the x-axis is constant then data file is treated as constant step size data.
>Why is it still better to use rebin instead of calculation_step
>(with which the mentioned interpolation does not happen)?
There would be no need to use rebin in cases where the XY/XYE daya is constant step size unless of course you want to reduce the number of data points. Also rebin is not a simple interpolation but rather a rebinning. If the rebinning step size is larger than the observed step size the effect of rebinning would be akin to collecting the data at less steps but with a larger receiving slit. In this way alll data is used in the rebinning and eds remain valid.
alan