urbanity.topology

High-performance network centrality computation using NetworKit, and utilities for merging NetworkX graph attributes into node GeoDataFrames.

from urbanity.topology import compute_centrality, merge_nx_attr, merge_nx_property

---

compute_centrality

compute_centrality(G, G_nodes, function, colname: str, T_nodes) -> GeoDataFrame

Computes a centrality measure for all nodes in the network using NetworKit's fast C++ backend, then merges the result back into the node GeoDataFrame.

Parameter	Type	Description
G	nx.MultiDiGraph	The street network graph.
G_nodes	GeoDataFrame	Node attribute GeoDataFrame to receive the centrality column.
function	callable	A NetworKit centrality class (e.g. networkit.centrality.Betweenness).
colname	str	Name of the output column added to G_nodes.
T_nodes	list	List of target node IDs for which centrality is computed.

Returns: GeoDataFrame with a new column colname containing centrality values.

Example:

import networkit as nk
from urbanity.topology import compute_centrality

nodes = compute_centrality(
    G, nodes,
    function=nk.centrality.Betweenness,
    colname="betweenness_centrality",
    T_nodes=list(G.nodes()),
)

Note

NetworKit converts the NetworkX graph internally. For very large networks (> 1M nodes), this step can take several minutes but is significantly faster than NetworkX's pure Python equivalents.

---

merge_nx_attr

merge_nx_attr(G, G_nodes, nxfunction, colname: str, T_nodes, temporal: bool = False) -> GeoDataFrame

Computes a NetworkX graph attribute function and merges the resulting values into the node GeoDataFrame.

Parameter	Type	Default	Description
G	nx.MultiDiGraph	required	Urban network graph.
G_nodes	GeoDataFrame	required	Node GeoDataFrame to augment.
nxfunction	callable	required	NetworkX function that returns a dict of {node_id: value}.
colname	str	required	Output column name.
T_nodes	list	required	Target node IDs.
temporal	bool	False	If True, appends a year suffix to colname.

Returns: Augmented GeoDataFrame.

Example:

import networkx as nx
from urbanity.topology import merge_nx_attr

nodes = merge_nx_attr(G, nodes, nx.pagerank, "pagerank", list(G.nodes()))

---

merge_nx_property

merge_nx_property(G_nodes, nxproperty, colname: str, T_nodes, temporal: bool = False) -> GeoDataFrame

Merges a pre-computed NetworkX graph property dictionary into the node GeoDataFrame. Use this when the property dict has already been computed (e.g. nx.betweenness_centrality(G)) to avoid recomputation.

Parameter	Type	Default	Description
G_nodes	GeoDataFrame	required	Node GeoDataFrame.
nxproperty	dict	required	Dict of {node_id: value} (output of a NetworkX function).
colname	str	required	Output column name.
T_nodes	list	required	Target node IDs to include.
temporal	bool	False	If True, appends a year suffix to colname.

Returns: Augmented GeoDataFrame.