Network Scores

Network Scores#

Network-topology scoring functions used by node-ranking workflows.

API path: node_ranking.network_scores

Topology-based score functions for node-ranking workflows.

The functions in this module operate directly on the network representation or on precomputed PTDF data. They do not require an economic-dispatch solution.

compute_edge_betweenness(G, weight=None, normalized=True)[source]#

Compute edge betweenness centrality for all edges.

Delegates to networkx.edge_betweenness_centrality with the provided weight and normalized options.

Parameters:

  • G (nx.Graph) – Input network.

  • weight (str or None, default=None) – Optional edge-attribute name used as a shortest-path weight. If None, the graph is treated as unweighted.

  • normalized (bool, default=True) – Whether to return normalized centrality values.

Returns:

Mapping from edge endpoints to edge-betweenness score.

Return type:

dict[tuple[hashable, hashable], float]

compute_node_betweenness_centrality(G)[source]#

Compute node betweenness centrality for all nodes.

Betweenness centrality is the fraction of shortest paths between node pairs that pass through each node. This implementation is unweighted and normalized.

Parameters:

G (nx.Graph) – Input network.

Returns:

Mapping from node label to normalized betweenness centrality.

Return type:

dict[hashable, float]

compute_node_degree_centrality(G)[source]#

Compute node degree centrality for all nodes.

Degree centrality is the node degree divided by the maximum possible degree, n - 1, so scores lie in [0, 1] for simple graphs.

Parameters:

G (nx.Graph) – Input network.

Returns:

Mapping from node label to normalized degree centrality.

Return type:

dict[hashable, float]

compute_node_ptdf_contribution_scores(
ptdf,
edges,
nodes,
mask,
G,
method='sum',
fmax_attr='F_max',
)[source]#

Compute PTDF-based node contribution scores.

Each non-slack node receives a score based on the magnitude of its PTDF column across all lines. The score can be aggregated by sum, by maximum line impact, or by a capacity-weighted sum. The slack node is assigned score 0 because PTDF columns are only defined for the non-slack buses in the reduced representation.

Parameters:

  • ptdf ((m, n-1) np.ndarray) – PTDF rows=lines, cols=non-slack buses (order = mask).

  • edges (list[(u, v, data)]) – Edges matching ptdf rows.

  • nodes (list) – All node labels (order used to build B).

  • mask (list[int]) – Indices of non-slack buses corresponding to ptdf columns.

  • G (nx.Graph) – Graph with edge attribute fmax_attr.

  • method ({"sum","max","weighted_sum"}) –

    • “sum”: sum_l abs(PTDF_{l,k})

    • ”max”: max_l abs(PTDF_{l,k})

    • ”weighted_sum”: sum_l abs(PTDF_{l,k}) * F_max(l)

  • fmax_attr (str) – Edge attribute used as weight for “weighted_sum”.

Returns:

scores – Contribution score for each node in the full node set. Slack node has score 0.

Return type:

dict[node_label, score]