iduedu.dijkstra_path_length_parallel¶

iduedu.dijkstra_path_length_parallel(urban_graph, *, source_nodes=None, gdf_sources=None, graph_node_column='graph_node_id', weight='time_min', cutoff=None, reverse=False, dtype=<class 'numpy.float32'>, max_workers=None)[source]¶

Run independent Dijkstra searches for each source.

Unlike multi_source_dijkstra_path_length(), sources are not merged into one queue. The result contains one sparse row per source object or source node, which makes this helper suitable for per-origin isochrone calculations.

Parameters:

urban_graph (UrbanGraph) – Urban graph with node and edge tables.
source_nodes (Optional[Iterable[Any]]) – Source node ids. Pass either this argument or gdf_sources.
gdf_sources (DataFrame | None) – DataFrame or GeoDataFrame with source objects.
graph_node_column (str) – Column containing graph node ids in gdf_sources.
weight (Literal['length_meter', 'time_min']) – Edge weight column.
cutoff (float | None) – Optional maximum path cost.
reverse (bool) – If True and the graph is directed, run on the reversed adjacency matrix.
dtype (dtype) – Floating dtype for returned sparse values.
max_workers (int | None) – Optional number of numba worker threads.

Return type:

DataFrame

Returns:

Sparse DataFrame whose rows are source objects and whose columns are reachable graph node ids. The normalized source mapping is stored in result.attrs["source_nodes"].