Adjacency matrix - IduEdu 1.2.1 documentation

# !pip install iduedu

# To read .parquet
# !pip install pyarrow

import geopandas as gpd

# Read a GeoDataFrame, for example this one
data = gpd.read_parquet('data/spb_buildings.parquet')

# Create a polygon that encircles all the buildings in the dataset to download the graph
polygon = data.to_crs(4326).union_all().convex_hull.buffer(0.001)

from iduedu import get_intermodal_graph, get_drive_graph, get_walk_graph

# Download the intermodal graph
G_intermodal = get_intermodal_graph(territory=polygon)

# Uncomment the following lines to download walking or driving graphs for the same area
# G_walk = get_walk_graph(polygon =polygon)
# G_drive = get_drive_graph(polygon =polygon)

2025-10-07 17:52:59.701 | INFO     | Downloading walk network via Overpass ...

2025-10-07 17:53:11.770 | WARNING  | Graph contains 58 weakly connected components. This means the graph has disconnected groups if edge directions are ignored. Component sizes:: [4235, 19, 9, 7, 5, …]
2025-10-07 17:53:11.782 | WARNING  | Removing 164 nodes from 57 smaller strongly connected components. These are subgraphs where nodes are internally reachable but isolated from the rest. Retaining only the largest strongly connected component (4235 nodes).

2025-10-07 17:54:32.280 | INFO     | Composing intermodal graph...
2025-10-07 17:54:33.021 | WARNING  | Removing 746 nodes from 359 smaller strongly connected components. These are subgraphs where nodes are internally reachable but isolated from the rest. Retaining only the largest strongly connected component (6062 nodes).

import numpy as np
from iduedu import get_adj_matrix_gdf_to_gdf

# Compute the adjacency matrix between the objects in the dataset using the intermodal graph.
# The weight of the edges is set to 'time_min' (travel time in minutes), and the data type is set to np.float32 for precision.
matrix = get_adj_matrix_gdf_to_gdf(data, data, G_intermodal, weight='time_min', dtype=np.float32)

# Uncomment the following lines to compute adjacency matrices using driving or walking graphs with different weights and data types
# matrix = get_adj_matrix_gdf_to_gdf(data,data,G_drive,weight='time_min',dtype=np.float16) # Use dtype >np.float16 for more precision
# matrix = get_adj_matrix_gdf_to_gdf(data,data,G_walk,weight='length_meter',dtype=np.float16) # Use dtype >np.float16 for more precision
matrix

	0	1	2	3	4	5	6	7	8	9	...	1248	1249	1250	1251	1252	1253	1254	1255	1256	1257
0	0.000000	19.477707	21.257921	19.482340	24.370579	22.620644	48.959641	21.764475	22.554434	19.405178	...	23.824213	14.421368	20.024340	20.100258	20.032118	27.937365	22.015039	20.085539	19.885014	22.431742
1	16.907707	0.000000	10.998884	6.203304	13.171542	11.171609	39.310604	12.825438	12.905397	8.206142	...	12.375177	7.542331	9.765304	9.151222	10.383081	16.738329	10.816002	8.606503	8.685979	11.232705
2	19.997921	10.858885	0.000000	6.273519	11.051758	14.601824	41.960819	16.255653	15.805614	6.086357	...	15.805391	10.632546	2.645518	12.581437	13.283297	13.978543	8.396217	10.486717	6.406193	9.602920
3	15.502339	6.203304	6.273519	0.000000	9.506178	10.106242	37.715240	11.760073	11.310033	4.540777	...	11.309811	6.136965	7.119937	8.085856	8.787716	13.072963	7.150637	5.531137	5.020613	7.567339
4	20.190578	11.051542	10.141758	8.366178	0.000000	14.794482	41.133480	16.448311	15.678271	7.159016	...	15.998051	10.825205	12.278176	12.774096	12.405954	15.361201	7.818875	9.999375	4.988852	8.575578
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1253	26.107365	16.968328	13.988544	14.282963	14.861201	20.711267	47.690262	22.365097	21.915058	12.675801	...	21.914837	16.741991	13.424962	18.690882	19.232740	0.000000	11.675661	16.596161	11.815638	14.892365
1254	18.705038	9.566002	8.396217	6.880637	9.608875	13.308942	39.317940	14.962771	13.862731	4.503475	...	14.512510	9.339664	9.032636	11.288555	10.590414	11.675661	0.000000	8.183835	3.173311	7.360038
1255	16.365540	7.226502	9.786717	4.541137	10.339375	10.969441	36.278435	12.623271	10.293231	5.713975	...	12.173010	7.000164	10.853136	8.949056	7.350914	14.916162	7.613835	0.000000	5.483811	8.510538
1256	16.715014	7.575979	6.406193	4.890613	7.088852	11.318917	36.487911	12.972747	11.032707	2.513451	...	12.522487	7.349640	8.542612	9.298532	7.760390	12.485638	3.173311	5.353811	0.000000	5.230014
1257	19.041740	9.902705	10.982920	7.217339	8.495578	13.645644	39.314640	15.299474	14.849434	7.040178	...	14.849213	9.676366	13.119339	11.625257	12.327117	15.462364	7.890038	9.530538	5.060014	0.000000

1258 rows × 1258 columns

Adjacency matrix¶