Re: Help w/fminunc compatibility

[Ben Abbott]

> On Nov 9, 2015, at 7:54 PM, Rik <address@hidden> wrote:
> 
> 11/9/15
> 
> Could someone with Matlab and the optimization toolbox run the code below
> and report back the results?
> 
> Thanks in advance,
> Rik
> 
> --Code--
> l=@(x) exp(sum(x.^2));
> [x,fval,info,output,grad,hess]=fminunc(l,[1 1])
> --End Code--
> 

Warning: Gradient must be provided for trust-region algorithm;
  using line-search algorithm instead. 
 In fminunc at 382 

Local minimum found.

Optimization completed because the size of the gradient is less than
the default value of the function tolerance.

<stopping criteria details>


Computing finite-difference Hessian using user-supplied objective function.

x =

     0     0


fval =

     1


info =

     1


output = 

       iterations: 1
        funcCount: 6
         stepsize: 0.0677
    firstorderopt: 1.4901e-08
        algorithm: 'medium-scale: Quasi-Newton line search'
          message: 'Local minimum found.

Optimization completed because the size of the gradient is less than
the def...'


grad =

   1.0e-07 *

    0.1490
    0.1490


hess =

   NaN   NaN
   NaN   NaN