Efficient way to have non negative elements in a matrix in CVX

Dear CVX Team,

I am solving a problem which constraints on its positive semidefinite matrix variable X of components x_{i,j} require that all of such components are non negative that is:

x_{i,j} \ge 0, \quad i=1,\dots,n,\quad j=1,\dots,n.

At the moment I am modelling these constraints with the following code

     for i=1:n
         for j=1:n

but knowing the inefficiencies that matlab has when it comes to for cycles, I was wondering if there is a better way to give these constraints in input to cvx.


V. L.

The following choices will work:
X(:) >= 0 (always works)
vec(X) >= 0 (always works, presuming you have the vec which comes with CVX)
X >= 0 (if not in sdp mode)

Thank you it works. It is a pity that it did not get any faster though.

V. L.