Network.Lstar

Network.Lstar(weight=None, flow=None, reduced: bool = False, method: paminco.linalg.InverseMethod = InverseMethod.CHOLESKY, safe: bool = True, **kwargs)

Pseudo-inverse of weighted Laplacian.

Parameters

weightnp.ndarray, optional

If None, weigths of Laplacian are determined based on flow.

flownp.ndarray, optional

Use flow to determine weights if weights is None.

reducedbool, default=False

Whether to return the reduced form, i.e., without first row and first column.

method: int, str, or InverseMethod, default=InverseMethod.CHOLESKY

Method used to calculate inverse.

safebool, default=True

If True (default), matrix will only be inverted if network is connected.

Returns

ndarray or CholeskyInverse

The pseudo inverse of the weighted Laplacian.

Raises

SingularLaplaceError

Network is not connected.

numpy.linalg.LinAlgError

Inverting matrix failed.

See also

Network.connected_components

Connectedness of the network.

Network.L

Laplacian matrix.

paminco.linalg.star_inv

generalized inverse of a matrix.

paminco.linalg.CholeskyInverse

Inverting a matrix using choleksy decomposition.