Hello.
I solved below problem using cvx and get the [- 0.0892, 0.0178] as a solution. but I know that the optimal solution is [1, -0.2].
How I can reach to optimal solution using cvx??
%%%% Problem%%%
g = [-32/31 -160/31];
A1 = [1 5];
n = length(g);
l = -ones(n,1);
u = ones(n,1);
cvx_begin
variable d(n)
minimize( gd )
subject to
A1d <= 0
l <= d <= u
cvx_end
g is proportional to A_1 so every point [x,y] satisfying x+5y=0 (and the bounds ) is an equally good optimal solution (gives the same optimal value). The optimal face is a segment. You cannot force the interior-point solver to converge to your preferred point in the optimal face.
Thank you very much.