Question: Applying constraints in Statistics[NonlinearFit](Matrix Form)

I have a regression problem for which I need to neglect a certain corner of parameter space because the function I am trying to fit for becomes imaginary in this region. I have everything written so that I may use NonlinearFitMatrixForm, which works really well when I hold one of the problematic parameters fixed and set the limits of the other to within the real region of parameter space. However, to extend this problem to fit for all parameters I need to tell the procedure to only look in the section of parameter space where some complicated function of two of the parameters is positive. Thanks in advance for any help, Daryl
Please Wait...