Date: Nov 21, 2012 7:51 PM
Author: Ludwig Greil
Subject: Problem with NonlinearModelFit

Hello everyone,

I have this problem fitting the Real part of a function that at some point
develops an imaginary part. My dataset is real-valued, and i try fitting
Re[f] and afterwards generating the MeanPredictionBands. The fitting of
the function goes fine, but as soon as i try to create the MeanPredictionBands, i get the following error:

Experimental`NumericalFunction::nnum: The function [...] is not a number at {fitparameters}.

Then I tried a normal function that is real-valued everywhere, and tried fitting Re[f] and then create the MeanPredictionBands and again, the same error appears.

The question I have is basically: Is there a way to do what i want to do?

Thank you all for your help,
Ludwig Greil