dagma.nonlinear.DagmaMLP.fc1_to_adj¶

dagma.nonlinear.DagmaMLP.fc1_to_adj() → numpy.ndarray¶

Computes the induced weighted adjacency matrix W from the first FC weights. Intuitively each edge weight \((i,j)\) is the L2 norm of the functional influence of variable i to variable j.

Returns:¶

\((d,d)\) weighted adjacency matrix

Return type:¶

np.ndarray