Trying to fit a curve with restrictions

[sleepyqueen]

I am trying to fit my experimental data with my own function y=a*exp(-x/b)+(1-a)*exp(-x/c) and I want the "a" paremeter to be >0 when doing the fitting.

Is there anyway to do so?

Thanks!

[altenbach]

Assuming you have a recent LabVIEW version, try the Constrained Nonlinear Curve Fit VI. Good luck!

Message Edited by altenbach on 04-03-2009 05:43 PM

[sleepyqueen]

what if I use 7.0 vision? Is there any way to realize Constrained Nonlinear Curve Fit use Labview 7.0?

[altenbach]

---

sleepyqueen wrote:
what if I use 7.0 vision? Is there any way to realize Constrained Nonlinear Curve Fit use Labview 7.0?

---

 Yes, but you have to do it all yourself from scratch. 😞

[DSPGuy]

Try fitting to (a^2)*exp(...) + (1-a^2)*exp(...)

a^2 is always positive, so the constraint is now implicit in the model.

-Jim