Info

This question is closed. Reopen it to edit or answer.

non linear fitting of two parameter, one parameter is never changing ...

1 view (last 30 days)
Hello, I am doing a non linear fit, my model is like this:
a = K * quad(@(x)(besselj(1.5,x)).^2 .* x./(x.^4 + t.^2) )
fitting parameters are "K" and "t". it is working fine, but it seems to fit only "t" and gives back "K" the same as I entered as the initial value. When I change K's initial value, the same value is given back but with different "t". Why this discrimination is happening?

Answers (1)

Daniel Armyr
Daniel Armyr on 5 Dec 2011
Assuming you use fminsearch, it will only optimize on the variable x. If you want to optimize on more than one variable, you have to make x a vector, and then replace x with x(1) and t with x(2).
If you read the lower part of the documentation for fminsearch, it will tell you this, but I myself allways mess up when doing this because I forget you can only have one variable to optimize on.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!