Error using fmincon, Supplied objective function must return a scalar value.

I have written following matlab code using fmincon function,

    x0=[0.5;0.5;]; %starting point
    Aieq=[1,1;repmat([(-eta * Pt * my_norm(f', H_DL(:,i))), (my_norm(W(:,i)', H_UL(:,i)))],K,1);];
    b=[1;repmat([0],K,1);]; %inequality constraint 
    Aeq=[]; beq=[];        %equality constraint 
    lb=[0;0;];ub=[1;1;];   %lower bound, upper bound 
    [tau,fun] = fmincon(@(tau) objecfun(tau,A,B,K,sigma_AP),x0,Aieq,b,Aeq,beq,lb,ub);

function [tau,fun] = objecfun(tau,A,B,K,sigma_AP)
   for i=1:K 
       fn(i)=-log10(1+(A(i)*tau(0+1))/((B(i)*tau(0+1)) + tau(1+1)* sigma_AP));
   end
   fun=sum(fn);
end

function [val]= my_norm(F,H)
   val=((sqrt(real(trace( (F * H)' * (F * H) ) ) ) )^2);
end

.

Please some one help me to solve the following error that I am obtaining while running the code.

Error using fmincon (line 639)
Supplied objective function must return a scalar value.

To run the above matlab code, when K=1, i can have only one value i=1, then parameter values can be taken as follows.

A=3.5153e+21;eta = 1;
H_DL =  1.0e+02 * [   7.3460 + 2.3079i;    2.0984 - 3.5173i;   4.7202 + 0.6043i;];
H_UL =  1.0e+02 * [   7.3460 + 2.3079i; 2.0984 - 3.5173i; 4.7202 + 0.6043i; ];
f =[ 1;  0;  0; ];
W =[ 1;  0;  0; ];
Pt =  1.0000e+10;
sigma_AP = 1; 

Answer 1

[Jon]

I'm not quite sure about the details of your problem but I think you want to do something closer to this.

Note the objective function should just return the objective function value for a given value of the design variable, in your case I think that is tau

    x0=[0.5;0.5;]; %starting point
    Aieq=[1,1;repmat([(-eta * Pt * my_norm(f', H_DL(:,i))), (my_norm(W(:,i)', H_UL(:,i)))],K,1);];
    b=[1;repmat([0],K,1);]; %inequality constraint 
    Aeq=[]; beq=[];        %equality constraint 
    lb=[0;0;];ub=[1;1;];   %lower bound, upper bound 
    [tau,fval] = fmincon(@(tau) objecfun(tau,A,B,K,sigma_AP),x0,Aieq,b,Aeq,beq,lb,ub);
function f= objecfun(tau,A,B,K,sigma_AP)
   for i=1:K 
       fn(i)=-log10(1+(A(i)*tau(0+1))/((B(i)*tau(0+1)) + tau(1+1)* sigma_AP));
   end
   f=sum(fn);
end