Hi all,

Need to fit data which corresponds to the amplitude and phase of an electrical signal as it goes through resonance. I could simply fit to the amplitude, throwing out the phase information, or fit to each data set separately, but it would obviously be best to fit to the whole data set with a single set of optimized parameters. I have never seen this addressed before: using one model to fit to multiple dependent variables.

My feeling is that I should be able to modify the chi-squared term to find the difference between the fitted function(s) and the actual data in a multi-dimensional space as a straight extension of the Levenberg-Marquardt method (but I haven't written the code yet). Has anyone ever done something like this?

Thanks,
Rob