Eigs function returns complex numbers when computing the largest eigenvalue

Hello, i am contructing an adjacency matrix from real networks as follows:

 x = dlmread('file.txt');
 left = x(:,1)+1;
 right = x(:,2)+1;
 Adj = sparse(left,right,1,7319,7319);
 eigs(Adj)

The networks are directed, some of them are very sparse while several others may also be disconnected. The problem is that for some networks i get the largest eigenvalue as a complex number. I have read several related threads to my post but i couldn't properly understand the quialitative reason.

Answer 1

[John D'Errico]

Your matrix is not symmetric. The eigenvalues of a non-symmetric matrix will often be complex numbers.

Answer 2

[Rupesh Rupesh]

Try using svd