Gaus exponential convolution

ROOT
1

Hi ,
I would like to fit a histogram with a gaussian convoluted with an exponential.
i had already a look in /fit/langaus.c but still not managed to modify it for my case.

I believe we have practically to change the part:
//=======
for(i=1.0; i<=np/2; i++)
{
xx = xlow + (i-.5) * step;
fland = TMath::Landau(xx,mpc,par[0]) / par[0];
sum += fland * TMath::Gaus(x[0],xx,par[3]);

     xx = xupp - (i-.5) * step;
     fland = TMath::Landau(xx,mpc,par[0]) / par[0];
     sum += fland * TMath::Gaus(x[0],xx,par[3]);
  }

//======

with the TMath::Exp(p0 + p1*xx)

but my different combinations did not worked out.

Has anyone gone through that?

thank you alot.

2

Hi,
I would compute the convolution of an exponential and the gaussian analytically using the error function ( or ROOT::Math::normal_cdf ) for the Gaussian integral.

Regards

Lorenzo