Error on integral of fit


I am plotting a a signal and background with a combined fit and then extracting the parameters to plot separate fits for the background and signal components, closely following the example: … emo.C.html

I need to find the integral of the background and signal parts along with their corresponding errors. I have no problem with the integral, but it is the calculation of the corresponding errors I am struggling with. How should I go about doing this?

I have read … alErrorand see that in order to use this i need pass the covariance matrix for the fit, however due to extracting the parameters with their errors from the combined fit I do not have a covariance matrix…

I hope this makes sense, thanks!


What is the problem exacting the covariance matrix ? You can get it from the TFitResult class (obtained fitting using option "S’).
You might need to use only a subset of the matrix for the relevant parameters.

Otherwise the alternative is to re-parametrize the functions and use the integral of the function a parameter of the fit.

Best Regards