Setting Value of Fit Parameter to Depend on Another Fit Parameter Value

Newbie
,
1

Hello all,

I would like to perform a fit, where the value of one of the fit parameters depends on the value of another fit parameter. For example, lets say I have a functions that has two parameters, then what I would like to do is the following:

Int_t fixedValue = 2;
fitFunction->SetParameter(0, 1.);
fitFunction->FixParameter(1, fitFunction->GetParamter(0) + fixedValue);

Basically, if I know the value of par[0], then par[1] = par[0] + fixedValue. However, this is a much more simplified version of my actual situation, so I do not want to define par[1] = par[0] + fixedValue in my user defined function, and I was wondering if I could obtain the same results by using the above block of code? Or, is there a better method to achieve what I want?

Thank you very much and any help is appreciated!

2

Hi @Bassam!

No, it does not work like this unfortunately. If you do fitFunction->FixParameter(1, fitFunction->GetParamter(0) + ...) it will take the value that the zeroth parameter currently has, it’s not that there two parameters are now connected.

Is what you’re doing a likelihood fit? If you want to do more complicated likelihood fits, then you should consider using RooFit. Is that an option for you?

Cheers,
Jonas

3

Thank you very much for your reply @Jonas! I will look into it and also try different ideas. Worst case scenario, I will resort to replacing par[1] with par[0] + fixedValue in the function definition.