Feasible Region and solvers

[Eduardo Mendes]

Hello

I am trying to get the feasible region for a LMi problem using three different solvers: lmilab, sedumi and mosek.  I am receiving a different answer for each one of the solvers.   For instance, when sedumi is used, the feasible region covers almost all the defined rectangle (10 by 10) whereas when lmilab is used, the feasible region is empty and for mosek a tiny region around zero.  Since I don't know what is the true feasible region, I am not sure if I choose tiny region - no region at all  or the whole thing.  No changes, except for the solver, were added to sdpsettings. 

Is there a guideline on how to choose the solver in such cases?  Is lmilab a solver choice under yalmip in such cases?   

Many thanks.

Ed

[Johan Löfberg]

I don't know what or how you are doing this, but the big difference between mosek and sedumi indicates that your problem is ill-posed somehow. lmilab you never use with yalmip

without reproducible code, nothing more can be said

[Eduardo Mendes]

Many thanks.