speigsh

• full name: tenpy.tools.math.speigsh

• parent module: tenpy.tools.math

• type: function

tenpy.tools.math.speigsh(A, k, *args, **kwargs)

Wrapper around scipy.sparse.linalg.eigsh(), lifting the restriction k < rank(A)-1.

Parameters:

• A (MxM ndarray or like scipy.sparse.linalg.LinearOperator) – The (square) hermitian linear operator for which the eigenvalues should be computed.

• k (int) – The number of eigenvalues to be computed.

• *args – Further arguments directly given to scipy.sparse.linalg.eigsh().

• **kwargs – Further keyword arguments directly given to scipy.sparse.linalg.eigsh().

Returns:

• w (ndarray) – Array of min(k, A.shape[0]) eigenvalues.

• v (ndarray) – Array of min(k, A.shape[0]) eigenvectors, v[:, i] is the i-th eigenvector. Only returned if kwargs['return_eigenvectors'] == True.