medil.evaluate.sfd

medil.evaluate.sfd(predicted_biadj, true_biadj)

Perform analysis of the distances between true and reconstructed structures :param biadj_mat: :type biadj_mat: input directed graph :param biadj_mat_recon: :type biadj_mat_recon: learned directed graph in the form of adjacency matrix

Returns:

• sfd (squared Frobenius distance (bipartite graph))

• ushd (structural hamming distance (undirected graph))