sf.apps.train.A_to_cov

A_to_cov(A)[source]

Convert an adjacency matrix to a covariance matrix of a GBS device.

The input adjacency matrix must have singular values not exceeding one. This can be achieved by rescaling an arbitrary adjacency matrix using rescale_adjacency().

Example usage:

>>> a = np.ones((3, 3))
>>> a = rescale_adjacency(a, 2, True)
>>> cov = A_to_cov(a)
Parameters

A (array) – the adjacency matrix

Returns

the covariance matrix of \(A\)

Return type

array