Create Hessian for fmincon from a large number of constraints
1 view (last 30 days)
Show older comments
Is there a way to create a Hessian with a variable number of constraint equations?
gradc = jacobian(c,x).';
hessc1 = jacobian(gradc(:,1))
hessc2 = jacobian(gradc(:,2))
hessc3 = jacobian(gradc(:,3))
hessc4 = jacobian(gradc(:,4))
hessch1 = matlabFunction(hessc1)
hessch2 = matlabFunction(hessc2)
hessch3 = matlabFunction(hessc3)
hessch4 = matlabFunction(hessc4)
myhess = @(x,lambda)(hessfh(x) + ...
lambda.ineqnonlin(1)*hessc1h(x) + ...
lambda.ineqnonlin(2)*hessc2h(x) + ...
lambda.ineqnonlin(3)*hessc3h(x) + ...
lambda.ineqnonlin(4)*hessc4h(x));
Above works for a fixed number of constraints. But, I can't figure out if it is possible to create the myhess handle with a variable number of constraints.
I tried the below code which I used a string to create the code that would create the myhess handle and converted it into a function handle, but I don't think fmincon with take the myhess handle because I had to include an extra input @(x,lambda,hessch)
gradc = jacobian(c,x).';
hessc = [];
for i=1:size(c,2)
hessc = [hessc;{jacobian(gradc(:,i) ,x);}];
end
hessch = [];
for i=1:size(c,2)
hessch = [hessch;{matlabFunction(hessc{i}, 'vars',{x})}];
end
myhess = '@(x,lambda,hessch)(hessfh(x)+';
for i=1:length(hessch)
myhess = [myhess,'lambda.ineqnonlin(',num2str(i),')*hessch{',num2str(i),'}(x)+'];
end
myhess(end)=')';
myhess = str2func(myhess)
I am not very familiar with the syntax of creating function handles and any help would be greatly appreciated.
0 Comments
Answers (0)
See Also
Categories
Find more on Nonlinear Optimization in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!