i have write the follwing equation but not solved
cvx_begin
variable Ges1(Tx,Tx) complex semidefinite
C=0;
for i=1:t
C=C+log(real((f(i) * (trace(Ges1 * cell{i})-thre(t)))));
end
maximize( log_det((Ges1))+C) %-2*log_det(S-Ges1)
cvx_end
I don’t know what you’re doing with cell{i} , but I’m not sure it’s allowed (but maybe it is). Does CVX accept your model? If so, what happens when you try to solve it?
I got rid of cell,chose Tx = 3;, used a random real symmetric 3 by 3 psd matrix for A, and made f and thre both ones(3,1). When I did that, SDPT3 failed, but SeDuMi succeeded. Note the warning from CVX about using the successive approximation method.
i try cell{i} in other code and i get result but with this code i get failed
What other code are you talking about? And what fails? Is CVX accepting your problem and the solver fails, or is CVX not accepting your problem?
CVX accept the problem but but the solver fail
Per the CVX User’s Guide:
… this is an experimental approach that works well in many cases, but cannot be guaranteed.
The bottom line, unfortunately, is that we cannot guarantee that the successive approximation approach will successfully handle your specific models. If you encounter problems, you are invited to submit a bug report, but we will not be able to promise a fix.
Have you tried other solvers? For instance, using the inputs I described previously, the problem was solved with SeDuMi, but failed with SDPT3.
I’m beginner to CVX if possible can i send you my simple code and figure out the problem
Show us the output from running it. Try all solvers you have
cvx_solver sedumi
cvx_solver sdpt3
cvx_solver mosek % if you have mosek capability
Successive approximation method to be employed.
For improved efficiency, SeDuMi is solving the dual problem.
SeDuMi will be called several times to refine the solution.
Original size: 272 variables, 92 equality constraints
3 exponentials add 24 variables, 15 equality constraints
Cones | Errors |
Mov/Act | Centering Exp cone Poly cone | Status
--------±--------------------------------±--------
3/ 3 | 8.000e+00 5.391e+01 0.000e+00 | Inaccurate/Solved
1/ 1 | 8.000e+00 8.282e+00 0.000e+00 | Inaccurate/Solved
1/ 1 | 7.903e+00s 1.335e-01 0.000e+00 | Solved
1/ 1 | 6.282e+00 2.178e-01s 0.000e+00 | Inaccurate/Solved
0/ 0 | 8.000e+00 0.000e+00 0.000e+00 | Inaccurate/Solved
0/ 0 | 8.000e+00 0.000e+00 0.000e+00 | Inaccurate/Solved
1/ 1 | 8.000e+00 1.605e-01 0.000e+00 | Inaccurate/Solved
1/ 1 | 7.323e+00s 2.456e-01s 0.000e+00 | Inaccurate/Solved
0/ 0 | 8.000e+00 0.000e+00 0.000e+00 | Inaccurate/Solved
Status: Inaccurate/Solved
Optimal value (cvx_optval): +28.75
Successive approximation method to be employed.
For improved efficiency, SDPT3 is solving the dual problem.
SDPT3 will be called several times to refine the solution.
Original size: 275 variables, 93 equality constraints
4 exponentials add 32 variables, 20 equality constraints
Cones | Errors |
Mov/Act | Centering Exp cone Poly cone | Status
--------±--------------------------------±--------
4/ 4 | 8.000e+00 1.927e+01 0.000e+00 | Failed
4/ 4 | 3.579e+00 6.972e-01 0.000e+00 | Failed
0/ 0 | 0.000e+00 0.000e+00 0.000e+00 | Failed
0/ 0 | 0.000e+00 0.000e+00 0.000e+00 | Failed
0/ 0 | 0.000e+00 0.000e+00 0.000e+00 | Failed
Status: Failed
Optimal value (cvx_optval): NaN
i have no mosek capability
Well, on the version (input data values) I solved, SeDuMi solved it (not Inaccurate/solved) and SDPT3 failed.
You should be happy SeDuMi even got Inaccurate solved. You will have to judge whether that is a useful solution for your purpose.
MOSEK has the reputation of being more numerically robust than SeDuMi or SDPT3. So perhaps if you can find someone with CVX/MOSEK capability (I don’t have it), it might do better than SeDuMi, but there is no guarantee that it will.