Is the new version compatible with MATLAB R2009a? And is there any simple example for test?

[Chao Yu]

Hi, I'm a new user of YALMIP, not quite familiar with it now. I install the tool box based on the directions of the YALMIP Wiki website.

I run yalmiptest, it works. Then I test the "Semidefinite Programming" of standard problems, 

http://users.isy.liu.se/johanl/yalmip/pmwiki.php?n=Tutorials.SemidefiniteProgramming

but I get a different result as the website shows.

The result on my matlab is "Marginal infeasibility: these LMI constraints may be feasible but are not strictly feasible"

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

|   ID|      Constraint|                  Type|   Primal residual|

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

|   #1|   Numeric value|     Matrix inequality|           0.23888|

|   #2|   Numeric value|     Matrix inequality|           0.88038|

|   #3|   Numeric value|   Equality constraint|      -2.7756e-016|

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

which the Primal residuals are different from the website. There is no random variables, so I think the values should be the same. 

Could you please indicate the reason? 

Many thanks

[Chao Yu]

Is it maybe because I install the toolbox wrongly? I also test some other examples, cannot get same results as the YALMIP website shows.

[Johan Löfberg]

R2009 should be no problems. Are you experiencing problems?

The LMI is solved perfectly fine. You can almost never expect strictly feasible solutions. To obtain that, you typically have to add explicit bounds such as X>=smallnumber*I. An error of 10^-16 is about as you ever ever could have in standard floating point numerics.

The message indicated "Marginal infeasibility: these LMI constraints may be feasible but are not strictly feasible" indicates to me that you are using LMILAB. Don't.

http://users.isy.liu.se/johanl/yalmip/pmwiki.php?n=Solvers.LMILAB

Install a better solver such as SeDuMi or SDPT3, or Mosek 7 for example.

[Chao Yu]

Thank you very much. I just install a SDPT3 toolbox from their website

http://www.math.nus.edu.sg/~mattohkc/sdpt3.html 

I run some examples on your website. The program works fine.

Thanks again :)