alan-jarvis
Hi,
Is there any documentation that describes the required format of TA data files?
I've looked in the Topas Tech Ref as well as the Users Manual and I don't see anything.
Specifically on step size and file-name extension.
I saw a vague reference in a NIST document that said you needed to rename TA files to "xye", but nothing about that in TA itself.
I binned my first set of data files at Diamond by 5:1 as well as 10:1 (0.001 re-binned to 0.005 and 0.010 degrees).
And the Diamond binning program renames them with an "xye" file-name extension.
TA runs the re-binned files OK, but not the raw data.
My raw data has a step size of 0.001 degrees.
TA returns an error "step size too small for my raw data step size of 0.001.
What is the minimum step size, and why is it that value?
I think you could start to lose some signal in the sharper peaks by binning it at some stage: I'd like to have the choice, and at least compare outputs.
An moderate sized peak reduces from 400 data points to 80 after re-binning at 5:1: still OK to fit.
But I can see that a very sharp peak might have its actual max intensity affected by re-binning.
Are there any statistical or computational reasons why we need to bin our data?
The only reason I can see is that computing time might be very long, however surely that's more a function of the model/number parameters than the step size.
Thanks,
Alan Jarvis
alancoelho
Alan
See section "10 FILE TYPES AND FORMATS" in the Technical_Reference.PDF
>Are there any statistical or computational
>reasons why we need to bin our data
Data does not have to be rebinned.
Also, there's no limit to how small step size can be. If there was an error saying step size too small then thats probably due to the RAW format of your data being different to what is expected.
cheers
alan
cerealreader
If you want to use the .dat files without rebinning, open the file in wordpad and remove the header. Then rename it to .xye file type. You should then be able to open the data in Topas without using the rebinned data.