Convert result of an optimisation problem back to original variables

1 次查看（过去 30 天）

Andrew Haselgrove 2019-5-30

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/464700-convert-result-of-an-optimisation-problem-back-to-original-variables

I have constructed an optimisation problem with the problem-based framework, but would like to apply a range of solvers, so use prob2struct to convert the problem to the solver-based framework. In the 'Algorithms' section, it describes that, for decision variables, this is is the same as calling vars = [x(:);y(:)] on the decision variables.

My question is, how can this best be reversed? I am currently using the following:

function sol = toProblemVars(x, prob)
    vars = prob.Variables;
    idx = prob.varindex();
    fields = fieldnames(vars);
    for i = 1:numel(fields)
        vars.(fields{i}) = zeros(size(vars.(fields{i})));
        vars.(fields{i})(:) = x(idx.(fields{i}));
    end
end