Hello,

I just could solve my problem. When I call the fit like this

Statistics:-NonlinearFit(ans, TVector, LVector, initialvalues=, output=parametervalues);

NonlinearFit seems to choose the first argument of ans as the independent one. I did it the other way around, just by accident. So now the initial value corresponds to the right variable in the model.

May this is an issue which could be mentioned in the help-file.

Many thanks anyway.

Physikuss