laplacian_centrality#
- laplacian_centrality(G, normalized=True, nodelist=None, weight='weight', walk_type=None, alpha=0.95)[source]#
Compute the Laplacian centrality for nodes in the graph
G
.The Laplacian Centrality of a node
i
is measured by the drop in the Laplacian Energy after deleting nodei
from the graph. The Laplacian Energy is the sum of the squared eigenvalues of a graph’s Laplacian matrix.\[ \begin{align}\begin{aligned}C_L(u_i,G) = \frac{(\Delta E)_i}{E_L (G)} = \frac{E_L (G)-E_L (G_i)}{E_L (G)}\\E_L (G) = \sum_{i=0}^n \lambda_i^2\end{aligned}\end{align} \]Where \(E_L (G)\) is the Laplacian energy of graph
G
, E_L (G_i) is the Laplacian energy of graphG
after deleting nodei
and \(\lambda_i\) are the eigenvalues ofG
’s Laplacian matrix. This formula shows the normalized value. Without normalization, the numerator on the right side is returned.- Parameters:
- Ggraph
A networkx graph
- normalizedbool (default = True)
If True the centrality score is scaled so the sum over all nodes is 1. If False the centrality score for each node is the drop in Laplacian energy when that node is removed.
- nodelistlist, optional (default = None)
The rows and columns are ordered according to the nodes in nodelist. If nodelist is None, then the ordering is produced by G.nodes().
- weight: string or None, optional (default=`weight`)
Optional parameter
weight
to compute the Laplacian matrix. The edge data key used to compute each value in the matrix. If None, then each edge has weight 1.- walk_typestring or None, optional (default=None)
Optional parameter
walk_type
used when callingdirected_laplacian_matrix
. One of"random"
,"lazy"
, or"pagerank"
. Ifwalk_type=None
(the default), then a value is selected according to the properties ofG
: -walk_type="random"
ifG
is strongly connected and aperiodic -walk_type="lazy"
ifG
is strongly connected but not aperiodic -walk_type="pagerank"
for all other cases.- alphareal (default = 0.95)
Optional parameter
alpha
used when callingdirected_laplacian_matrix
. (1 - alpha) is the teleportation probability used with pagerank.
- Returns:
- nodesdictionary
Dictionary of nodes with Laplacian centrality as the value.
- Raises:
- NetworkXPointlessConcept
If the graph
G
is the null graph.- ZeroDivisionError
If the graph
G
has no edges (is empty) and normalization is requested.
Notes
The algorithm is implemented based on [1] with an extension to directed graphs using the
directed_laplacian_matrix
function.References
[1]Qi, X., Fuller, E., Wu, Q., Wu, Y., and Zhang, C.-Q. (2012). Laplacian centrality: A new centrality measure for weighted networks. Information Sciences, 194:240-253. https://math.wvu.edu/~cqzhang/Publication-files/my-paper/INS-2012-Laplacian-W.pdf
Examples
>>> G = nx.Graph() >>> edges = [(0, 1, 4), (0, 2, 2), (2, 1, 1), (1, 3, 2), (1, 4, 2), (4, 5, 1)] >>> G.add_weighted_edges_from(edges) >>> sorted((v, f"{c:0.2f}") for v, c in laplacian_centrality(G).items()) [(0, '0.70'), (1, '0.90'), (2, '0.28'), (3, '0.22'), (4, '0.26'), (5, '0.04')]