MatrixWeighting.matrix_decomp

MatrixWeighting.matrix_decomp(cache=None)[source]

Compute a Hermitian eigenbasis decomposition of the matrix.

Parameters

cachebool or None, optional: If True, store the decomposition internally. For None, the cache_mat_decomp from class initialization is used.

Returns

eigvalnumpy.ndarray: One-dimensional array of eigenvalues. Its length is equal to the number of matrix rows.
eigvecnumpy.ndarray: Two-dimensional array of eigenvectors. It has the same shape as the decomposed matrix.

See Also

scipy.linalg.decomp.eigh :: Implementation of the decomposition. Standard parameters are used here.

Raises

NotImplementedError: if the matrix is sparse (not supported by scipy 0.17)