Pdf and function with different normalization (RooWrapperPdf)

[Ludwig]

Hi,

I created a personal pdf with RooWrapperPdf and my C++ Code.
In order to use the parameters in ongoing calculations, I wanted to know how I can get the right normalization factor for the function/parameters. I noticed when I use the same parameters on the same function, then the function doesn’t look the same as in the RooPlot frame (different normalization):

Blue fit is the pdf in RooPlot and red fit the same function as TF1.
My function is a simple gaussian with an amplitue parameter added:

[[nodiscard]] Double_t my_gauss(const Double_t x, const Double_t par) {
return par[0] * exp(-(x[0]-par[1])(x[0]-par[1])/(2par[2]*par[2]));
}

I’ve read different approaches to this problem in the forum, but those haven’t worked for me. Does anyone have a “simple” solution to this problem that I haven’t discovered yet?

Thanks in advance
Ludwig

[bellenot]

Maybe @jonas can help with this

[Ludwig]

I noticed that the amplitude parameter has a very big error value but there is no sign of any correlation between parameters. I guess the problem originates from there. But how do I fix that?