Deconvolution question

Hi, I have a distribution which is a product of a gaussian and a number of other lesser functions.

I already know the parameters for the gaussian part and wish to deconvolve this from the distribution.

An idea I had was to fit a custom function (a sum of functions) to the distribution which included a gaussian component with fixed parameters. Although this would give me the contributions from the other parts of the custom function, it would not allow me to draw the deconvoluted data, only the functions themselves (which is not ideal).

Does anyone have any ideas how to deconvolute a known gaussian from a distribution that is not purely gaussian?

Cheers, Ben.

Attached is what I was trying to do.

The result is to take a distribution (gaussian in this case, but can be any) and remove the contribution of a gaussian from the distribution.

div.C is an extra macro which outputs code to go into drawgaus.C -> faster than typing.

Hope this helps someone. Ben.
div.C (486 Bytes)
drawgaus.C (11.6 KB)

better code…
deconve.C (1.95 KB)
testhists.root (7.69 KB)