Optimal value is NaN, after 3th iteration of CVX

[Kevin3 (Kevin)]

@Mark_L_Stone Hi Mark, I am solving a SCA problem, after 3th iteration of CVX, the optimal value is NaN, what the reasons of this, and how to handle this. Thank you.

after 2th iteration

after 3th iteration

[Mark_L_Stone]

The result obtained using the SeDuMi solver is inaccurate. The optimal solution is not as good as before the optimization.

Unsafeguarded (no line search or trust region)Successive Convex Approximation (SCA) or generally, alternating variables optimization are unreliable. It might not descend (for minimization problem), i.e., iterates could get worse. It might not converge to anything; and if it does converge, it might not be to a local optimum of the original problem, let alone a global optimum. The solution of successive iterations, and therefore subproblem inputs, can become wilder and wilder, until at some point …

[Kevin3 (Kevin)]

Mark, my solver is mosek

[Mark_L_Stone]

Same answer. Perhaps Mosek makes it to a later iteration before the problem fails.

[Kevin3 (Kevin)]

and … what i shou do next

[Mark_L_Stone]

You can continue on your path, play around with things, and try to improve the SCA algorithm; or just use a non-convex solver, perhaps under YALMIP. Only you know thee reason why you’re trying to solve this problem using SCA, or trying to solve it at all. The goal of this forum is to help people solve convex optimization problems with CVX; the goal is not to ensure that whatever paper someone tries to replicate will lead to a publishable research paper or thesis.

Why isn't CVX accepting my model? READ THIS FIRST!

[Kevin3 (Kevin)]

Thank you very much Mark.