Thread Subject: Passing constants to optimization functions

I am trying to use lsqnonlin(), but my user defined function
needs to know a couple of constants. The lsqnonlin expects
my function to only accept one parameter (the one being
optimized). Is there a better way to "pass" these constants
besides making them global?

Thanks,

David

"David Doria" <daviddoria@gmail.com> wrote in message <feiqhl$7nd
$1@fred.mathworks.com>...
> I am trying to use lsqnonlin(), but my user defined function
> needs to know a couple of constants. The lsqnonlin expects
> my function to only accept one parameter (the one being
> optimized). Is there a better way to "pass" these constants
> besides making them global?

Use an anonymous function.

a = 2;
f = @(x) (x-a).^2;

Obviously, the minimum occurs at x = a = 2.
Can we find it?

x = fminsearch(f,0)
x =
                         2

HTH,
John