I keep getting this error (Failure in initial user-supplied objective function evaluation. FSOLVE cannot continue)

My currenct script looks like this:

 function F = myfun(x)

      x0 = [0.1 0.2 0.3 0.4]
      x = x0
      f1 = inline('1*(1 - x(1))+100*(-1*x(1)*x(2))');
      f2 = inline('1*(2 - x(2))+100*(-1*x(1)*x(2)-1*x(2)*x(3))');
      f3 = inline('1*(0 - x(3))+100*(1*x(1)*x(2)-1*x(2)*x(3))');
      f4 = inline('1*(0 - x(4))+100*(1*x(2)*x(3))');
      F = [f1(x);f2(x);f3(x);f4(x)]

      options=optimset('Display','iter');
      x = fsolve(@myfun,x0);

end

Still nothing seems to get this working, can anyone help me out? If i do not define my first [x] he will not run because there are not enough input arguments.

Thanks in advance!

Frenk

Answer 1

[Matt J]

The call to fsolve and the definition of myfun should take place in two separate workspaces. The function definition should look like this,

    function F = myfun(x)

      f1 = 1*(1 - x(1))+100*(-1*x(1)*x(2));
      f2 = 1*(2 - x(2))+100*(-1*x(1)*x(2)-1*x(2)*x(3));
      f3 = 1*(0 - x(3))+100*(1*x(1)*x(2)-1*x(2)*x(3));
      f4 = 1*(0 - x(4))+100*(1*x(2)*x(3));
      F = [f1;f2;f3;f4];

    end

The call to fsolve should be in a different function and look like this,

      x0 = [0.1 0.2 0.3 0.4];
      options=optimset('Display','iter');
      x = fsolve(@myfun,x0,options);

Notice also that there is no need to be using inline() in this situation. Moreover, unless you have a really old version of MATLAB, you should be discontinuing the use of inline() altogether and using Anonymous Functions instead.

[Matt J]

Frenk's Comment:

Thank you for your explanation, the question has been solved and i can continue with my proceedings. I will keep in mind not using the inline command.

Answer 2

[Yahia Mounir]

Hello I am trying to run the following file however it give me error undefined variable X, I attached the my file can you help with that ,

Please note if i call the function at command window with the following steps @myfun t x0=[0 0] x=(@myfun,xo) it is working and giving me the answer i need only to use the mfile to get the answer

[Walter Roberson]

Just like Matt J explained, you need to be using two different files.

File myfun.m

function F = myfun(x)
   F = [2*x(1) - x(2) - exp(-x(1));
      -x(1) + 2*x(2) - exp(-x(2))];
end

file myfun_driver.m

x0=[0 0];
options = optimoptions('fsolve', 'Display', 'iter'); % Option to display output
x = fsolve(@myfun,x0,options);

Then you would invoke

myfun_driver

to run the code.

Answer 3

[Henn]

My current script looks like this:

 for kk=1:+1:21;
    sigma=sigmastart+0.01*(kk-1)
    beta1 = 0.5-alpha/sigma^2+sqrt((0.5-alpha/sigma^2)^2+2*rho/sigma^2);
    beta2 = 0.5-alpha/sigma^2-sqrt((0.5-alpha/sigma^2)^2+2*rho/sigma^2);
    e=0:+0.001:100;
    %guesses for investment and switching thresholds
    xi010=(beta1/(beta1-1))*(rho-alpha)*(H*F*rho*I)/rho;
    xi120=(beta2/(beta2-1))*(rho-alpha)*(H*F/rho-S12);
    xi210=(beta1/(beta1-1))*(rho-alpha)*(H*F/rho+S21);
    a010=xi010^(1-beta1)/(beta1*(rho-alpha));
    a120=-xi120^(1-beta2)/(beta2*(rho-alpha));
    a210=xi210^(1-beta1)/(beta1*(rho-alpha));
    %initial guess for operational decisions
    vars0 = [a120 a210 xi120 xi210];
    %specify options for fsolve
    opts=optimset('fsolve');
    opts=optimset(opts , 'Maxiter' , 2000 , 'Tolx' , 1e-6, 'tolfun' , 1e-6);
vars = fsolve(@fffRoinvof , vars0 , opts , beta1 , beta2 , alpha , H , F , rho , S12, S21);

and my function file looks like this

 function f = fffROinvof (vars,beta1,beta2,alpha,H,F,rho,S12,S21)
 a12=vars (1);
a21=vars (2);
xi12=vars (3);
xi21=vars (4);
f=zeros(4,1);
f(1)=a12*xi12^beta2+xi12/(rho-alpha)- H*F/rho-a21*xi12^beta1+S12;
f(2)=beta2*ai12*xi12^(beta1-1);
f(3)=-a21*xi21^beta1+...
    1/(rho-alpha)-H*F/rho+a12*xi21^beta2-S21;
f(4)=-beta1*a21*xi21^(beta1-1)+...
    1/(rho-alpha)+beta2*a12*xi21^(beta2-1);
end

However, I still get the error message when I run my script Error in Code (line 34) vars = fsolve(@fffRoinvof , vars0 , opts , beta1 , beta2 , alpha , H , F , rho , S12, S21);

Caused by: Failure in initial objective function evaluation. FSOLVE cannot continue.

[Walter Roberson]

Change

vars = fsolve(@fffRoinvof , vars0 , opts , beta1 , beta2 , alpha , H , F , rho , S12, S21);

to

vars = fsolve(@(vars) fffRoinvof(vars, beta1 , beta2 , alpha , H , F , rho , S12, S21), vars0 , opts );