Hi, I’ve written some code that utilizes the deconvolution feature in TSpectrum.
My questions relate to the input parameter used by the algorithm.
pfinder.Deconvolution(source,resp,size,X,Y,Z);
X - number of iterations. So, the higher this is, the finer the detail of the distribution taken into account.
Y - repetitions. The number of times the algorithm processes the input. Does this deconvolve the response from the source Y times? Or something to do with the algorithm converging?
Z - ‘boost’ - which I have yet to discern any effect.
I attach working code +input and pictures which show the effect of the parameters X.Y.Z.output.png
If anyone has advice on the optimum values for the parameter it would be appreciated.
Cheers, Ben.
[incidentally, directing me to Miroslavs paper isn’t ideal]
deconve.C (1.48 KB)
Shift.root (4.23 KB)
