Hex2vec embedder

In [1]:

from srai.embedders import Hex2VecEmbedder
from srai.joiners import IntersectionJoiner
from srai.loaders import OSMOnlineLoader
from srai.neighbourhoods import H3Neighbourhood
from srai.regionalizers import H3Regionalizer, geocode_to_region_gdf
from srai.plotting import plot_regions, plot_numeric_data
from pytorch_lightning import seed_everything

from srai.embedders import Hex2VecEmbedder from srai.joiners import IntersectionJoiner from srai.loaders import OSMOnlineLoader from srai.neighbourhoods import H3Neighbourhood from srai.regionalizers import H3Regionalizer, geocode_to_region_gdf from srai.plotting import plot_regions, plot_numeric_data from pytorch_lightning import seed_everything

In [2]:

SEED = 71
seed_everything(SEED)

SEED = 71 seed_everything(SEED)

Seed set to 71

Out[2]:

71

Load data from OSM¶

First use geocoding to get the area

In [3]:

area_gdf = geocode_to_region_gdf("Wrocław, Poland")
plot_regions(area_gdf, tiles_style="CartoDB positron")

area_gdf = geocode_to_region_gdf("Wrocław, Poland") plot_regions(area_gdf, tiles_style="CartoDB positron")

Out[3]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Next, download the data for the selected region and the specified tags. We're using OSMOnlineLoader here, as it's faster for low numbers of tags. In a real life scenario with more tags, you would likely want to use the OSMPbfLoader.

In [4]:

tags = {
    "leisure": "park",
    "landuse": "forest",
    "amenity": ["bar", "restaurant", "cafe"],
    "water": "river",
    "sport": "soccer",
}
loader = OSMOnlineLoader()

features_gdf = loader.load(area_gdf, tags)

folium_map = plot_regions(area_gdf, colormap=["rgba(0,0,0,0)"], tiles_style="CartoDB positron")
features_gdf.explore(m=folium_map)

tags = { "leisure": "park", "landuse": "forest", "amenity": ["bar", "restaurant", "cafe"], "water": "river", "sport": "soccer", } loader = OSMOnlineLoader() features_gdf = loader.load(area_gdf, tags) folium_map = plot_regions(area_gdf, colormap=["rgba(0,0,0,0)"], tiles_style="CartoDB positron") features_gdf.explore(m=folium_map)

Downloading sport: soccer      : 100%|██████████| 7/7 [00:02<00:00,  3.00it/s]

Out[4]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Prepare the data for embedding¶

After downloading the data, we need to prepare it for embedding. Namely - we need to regionalize the selected area, and join the features with regions.

In [5]:

regionalizer = H3Regionalizer(resolution=9)
regions_gdf = regionalizer.transform(area_gdf)
plot_regions(regions_gdf, tiles_style="CartoDB positron")

regionalizer = H3Regionalizer(resolution=9) regions_gdf = regionalizer.transform(area_gdf) plot_regions(regions_gdf, tiles_style="CartoDB positron")

Out[5]:

Make this Notebook Trusted to load map: File -> Trust Notebook

In [6]:

joiner = IntersectionJoiner()
joint_gdf = joiner.transform(regions_gdf, features_gdf)
joint_gdf

joiner = IntersectionJoiner() joint_gdf = joiner.transform(regions_gdf, features_gdf) joint_gdf

Out[6]:

region_id	feature_id
891e204756fffff	way/342560157
891e20471d7ffff	way/342560157
891e20471d3ffff	way/342560157
891e204716fffff	relation/3674824
891e204717bffff	relation/3674824
...	...
891e2042a6fffff	way/1044720759
	node/1984033864
	way/310477092
	way/1224586685
891e20406c7ffff	way/313424655

3949 rows × 0 columns

Embedding¶

After preparing the data we can proceed with generating embeddings for the regions.

In [7]:

import warnings

neighbourhood = H3Neighbourhood(regions_gdf)
embedder = Hex2VecEmbedder([15, 10])

with warnings.catch_warnings():
    warnings.simplefilter("ignore")
    embeddings = embedder.fit_transform(
        regions_gdf,
        features_gdf,
        joint_gdf,
        neighbourhood,
        trainer_kwargs={"max_epochs": 5, "accelerator": "cpu"},
        batch_size=100,
    )
embeddings

import warnings neighbourhood = H3Neighbourhood(regions_gdf) embedder = Hex2VecEmbedder([15, 10]) with warnings.catch_warnings(): warnings.simplefilter("ignore") embeddings = embedder.fit_transform( regions_gdf, features_gdf, joint_gdf, neighbourhood, trainer_kwargs={"max_epochs": 5, "accelerator": "cpu"}, batch_size=100, ) embeddings

100%|██████████| 3168/3168 [00:00<00:00, 31264.61it/s]
GPU available: False, used: False
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs

  | Name    | Type       | Params
---------------------------------------
0 | encoder | Sequential | 280   
---------------------------------------
280       Trainable params
0         Non-trainable params
280       Total params
0.001     Total estimated model params size (MB)

`Trainer.fit` stopped: `max_epochs=5` reached.

Out[7]:

	0	1	2	3	4	5	6	7	8	9
region_id
891e2040173ffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045
891e204756fffff	0.096392	0.220119	0.465636	0.581243	-0.187880	-0.037154	0.366241	-0.418458	0.386511	0.119130
891e204716fffff	-0.330573	0.020956	-0.333520	-0.457863	0.017288	0.245358	-0.368664	0.331725	-0.074456	-0.032008
891e20414b3ffff	-0.330574	0.020956	-0.333520	-0.457863	0.017288	0.245358	-0.368664	0.331725	-0.074456	-0.032008
891e20442d3ffff	-0.464493	0.138147	-0.393941	-0.574460	-0.067107	0.280365	-0.315895	0.398935	-0.085970	-0.146017
...	...	...	...	...	...	...	...	...	...	...
891e20406c7ffff	-0.596529	0.255250	-0.453033	-0.690959	-0.152528	0.314501	-0.261384	0.465135	-0.097409	-0.261858
891e2040d9bffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045
891e204e1d3ffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045
891e205a9a7ffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045
891e2042e7bffff	-0.207339	-0.090883	-0.147170	-0.687353	-0.153512	0.164471	0.044825	0.309862	-0.101235	-0.483526

3168 rows × 10 columns

Visualizing the embeddings' similarity¶

In [8]:

from sklearn.cluster import KMeans

clusterizer = KMeans(n_clusters=5, random_state=SEED)
clusterizer.fit(embeddings)

embeddings["cluster"] = clusterizer.labels_
embeddings

from sklearn.cluster import KMeans clusterizer = KMeans(n_clusters=5, random_state=SEED) clusterizer.fit(embeddings) embeddings["cluster"] = clusterizer.labels_ embeddings

/opt/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning
  super()._check_params_vs_input(X, default_n_init=10)

Out[8]:

	0	1	2	3	4	5	6	7	8	9	cluster
region_id
891e2040173ffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045	1
891e204756fffff	0.096392	0.220119	0.465636	0.581243	-0.187880	-0.037154	0.366241	-0.418458	0.386511	0.119130	3
891e204716fffff	-0.330573	0.020956	-0.333520	-0.457863	0.017288	0.245358	-0.368664	0.331725	-0.074456	-0.032008	2
891e20414b3ffff	-0.330574	0.020956	-0.333520	-0.457863	0.017288	0.245358	-0.368664	0.331725	-0.074456	-0.032008	2
891e20442d3ffff	-0.464493	0.138147	-0.393941	-0.574460	-0.067107	0.280365	-0.315895	0.398935	-0.085970	-0.146017	2
...	...	...	...	...	...	...	...	...	...	...	...
891e20406c7ffff	-0.596529	0.255250	-0.453033	-0.690959	-0.152528	0.314501	-0.261384	0.465135	-0.097409	-0.261858	2
891e2040d9bffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045	1
891e204e1d3ffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045	1
891e205a9a7ffff	0.343880	-0.223290	-0.012038	0.266015	0.214774	-0.123038	-0.256550	-0.107333	-0.217049	0.302045	1
891e2042e7bffff	-0.207339	-0.090883	-0.147170	-0.687353	-0.153512	0.164471	0.044825	0.309862	-0.101235	-0.483526	2

3168 rows × 11 columns

In [9]:

plot_numeric_data(regions_gdf, "cluster", embeddings)

plot_numeric_data(regions_gdf, "cluster", embeddings)

Out[9]:

Make this Notebook Trusted to load map: File -> Trust Notebook