Adjacency relations

import pandas as pd

blocks_gdf = pd.read_pickle('./../data/saint_petersburg/blocks.pickle')
blocks_gdf.head(3)

	geometry	site_area	residential	business	recreation	industrial	transport	special	agriculture	land_use	...	capacity_prison	count_prison	capacity_landfill	count_landfill	capacity_plant_nursery	count_plant_nursery	capacity_greenhouse_complex	count_greenhouse_complex	capacity_warehouse	count_warehouse
0	POLYGON ((349424.859 6631180.891, 349424.751 6...	0.032316	0.000	0.0	0.000000	0.0	1.000000	0.0	0.000000	LandUse.TRANSPORT	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
1	POLYGON ((352083.617 6633950.146, 352240.448 6...	137101.088247	0.099	0.0	0.079912	0.0	0.401072	0.0	0.417018	LandUse.AGRICULTURE	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
2	POLYGON ((346700.642 6618453.176, 346681.107 6...	95239.084434	1.000	0.0	0.000000	0.0	0.000000	0.0	0.000000	LandUse.RESIDENTIAL	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

3 rows × 137 columns

Adjacency graph

Generate adjacency graph indicating relations between adjacent blocks.

from blocksnet.relations import generate_adjacency_graph

adjacency_graph = generate_adjacency_graph(blocks_gdf, 10)

2025-09-19 19:37:40.874 | INFO     | blocksnet.relations.adjacency.core:_generate_adjacency_nodes:9 - Generating nodes
2025-09-19 19:37:40.880 | INFO     | blocksnet.relations.adjacency.core:_generate_adjacency_edges:14 - Generating edges
2025-09-19 19:37:43.924 | SUCCESS  | blocksnet.relations.adjacency.core:generate_adjacency_graph:37 - Adjacency graph successfully generated: 9533 nodes, 30567 edges

Now adjacency_graph is a networkx.Graph instance and can be operated using its methods. For example, one can get block_id=2 neighbors using g.neighbors(2).

block_id = 2

neighbors = adjacency_graph.neighbors(block_id)
block_gdf = blocks_gdf.loc[[block_id]]
neighbors_gdf = blocks_gdf.loc[neighbors]

ax = block_gdf.plot(color='red', figsize=(6,6), edgecolor='black')
neighbors_gdf.plot(ax=ax, color='blue', edgecolor='black')
ax.set_title(f'block_id={block_id} neighbors')
ax.set_axis_off()

Adjacency context

Get adjacency graph’s subgraph for blocks provided.

from blocksnet.relations import get_adjacency_context

context_graph = get_adjacency_context(adjacency_graph, block_gdf)
len(context_graph.nodes), len(context_graph.edges)

(9, 17)