Determinant Maximization problem

[Sandeep hegde]

Hello members, I am trying to solve the 'log_det' problem of the form

Max det(Y)

subject to [[Y,  (A-BL)'Y],

                   [Y(A-BL),  Y]] >=0

L_i@Y@L_i' <=1

Here A,B, and L are numpy arrays and values are known. Y is cvx variable of shape (n,n)

Here is my code:

Y = cvx.Variable((n,n))

obj = cvx.Maximize(cvx.log_det(Y))

constraints = []

F = np.array([

[Y, Y@(A-B@L).T],

[(A-B@L)@Y, Y ]])

for i in range(m):

   constraints+= [L[i,:]@Y@L[i,:].T<=1]

 constraints+= [F>>0]      //F should be positive semidefinite

p = cvx.Problem(obj,constraints)

res = p.solve()    //error showing line

When I use this function, I get the error:

AttributeError: 'numpy.ndarray' object has no attribute 'parameters'.

What am I missing in the formulation? Can someone please help me out?

Thank you

Sandeep

[Steven Diamond]

It should be 

```

F = cvx.bmat([

[Y, Y@(A-B@L).T],

[(A-B@L)@Y, Y ]])

```

Play around with that.