Hex2vec embedder

Run this notebook in Google Colab:

Remember to install the srai library before running the notebook:

%pip install srai[all]

In [1]:

Copied!





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.plotting import plot_numeric_data, plot_regions
from srai.regionalizers import H3Regionalizer, geocode_to_region_gdf
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.plotting import plot_numeric_data, plot_regions
from srai.regionalizers import H3Regionalizer, geocode_to_region_gdf

In [2]:

Copied!

SEED = 71
seed_everything(SEED)
SEED = 71
seed_everything(SEED)

Seed set to 71

Out[2]:

Load data from OSM¶

First use geocoding to get the area

In [3]:

Copied!

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]:

Copied!





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)

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]:

Copied!

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]:

Copied!

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
891e2040897ffff	node/280727473
891e2040d4bffff	node/300461026
891e2040d4bffff	node/300461036
891e2040d5bffff	node/300461042
891e2040887ffff	node/300461045
...	...
891e2042637ffff	way/1360073315
891e20420cbffff	way/1360073315
891e20420dbffff	way/1360073315
891e20420c3ffff	way/1360073315
891e2040817ffff	way/1366196846

4065 rows × 0 columns

Embedding¶

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

In [7]:

Copied!





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

GPU available: False, used: False

TPU available: False, using: 0 TPU cores

HPU available: False, using: 0 HPUs

  | Name    | Type       | Params | Mode 
-----------------------------------------------
0 | encoder | Sequential | 280    | train
-----------------------------------------------
280       Trainable params
0         Non-trainable params
280       Total params
0.001     Total estimated model params size (MB)
4         Modules in train mode
0         Modules in eval mode

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

Out[7]:

	0	1	2	3	4	5	6	7	8	9
region_id
891e204050fffff	0.095495	0.138996	0.566001	0.688587	-0.025211	-0.170529	0.644415	-0.266062	0.530560	-0.066391
891e20434d3ffff	0.338901	-0.193323	-0.014221	0.313743	0.227657	-0.132642	-0.284652	-0.104796	-0.200931	0.378753
891e2051973ffff	-0.310405	0.045352	-0.296344	-0.414064	-0.021333	0.244782	-0.422900	0.357396	-0.103730	0.055133
891e2041497ffff	-0.310405	0.045352	-0.296344	-0.414065	-0.021333	0.244782	-0.422900	0.357396	-0.103730	0.055133
891e204083bffff	1.038319	-0.183254	0.187057	0.140822	0.123709	-0.830628	1.430029	-0.910638	0.197307	-0.844142
...	...	...	...	...	...	...	...	...	...	...
891e20404c3ffff	0.338901	-0.193323	-0.014221	0.313743	0.227657	-0.132642	-0.284652	-0.104796	-0.200931	0.378753
891e20474afffff	0.338901	-0.193323	-0.014221	0.313743	0.227657	-0.132642	-0.284652	-0.104796	-0.200931	0.378753
891e2042323ffff	-0.578869	0.324108	-0.507577	-0.754409	-0.084290	0.240572	-0.252477	0.415875	-0.184438	-0.246741
891e2051873ffff	-0.310405	0.045352	-0.296344	-0.414064	-0.021333	0.244782	-0.422900	0.357396	-0.103730	0.055133
891e2042e2bffff	-0.785091	0.567634	0.190860	-0.089924	-0.219985	0.231418	0.420068	0.015617	0.521316	-0.284058

3168 rows × 10 columns

Visualizing the embeddings' similarity¶

In [8]:

Copied!

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

Out[8]:

	0	1	2	3	4	5	6	7	8	9	cluster
region_id
891e204050fffff	0.095495	0.138996	0.566001	0.688587	-0.025211	-0.170529	0.644415	-0.266062	0.530560	-0.066391	3
891e20434d3ffff	0.338901	-0.193323	-0.014221	0.313743	0.227657	-0.132642	-0.284652	-0.104796	-0.200931	0.378753	0
891e2051973ffff	-0.310405	0.045352	-0.296344	-0.414064	-0.021333	0.244782	-0.422900	0.357396	-0.103730	0.055133	1
891e2041497ffff	-0.310405	0.045352	-0.296344	-0.414065	-0.021333	0.244782	-0.422900	0.357396	-0.103730	0.055133	1
891e204083bffff	1.038319	-0.183254	0.187057	0.140822	0.123709	-0.830628	1.430029	-0.910638	0.197307	-0.844142	2
...	...	...	...	...	...	...	...	...	...	...	...
891e20404c3ffff	0.338901	-0.193323	-0.014221	0.313743	0.227657	-0.132642	-0.284652	-0.104796	-0.200931	0.378753	0
891e20474afffff	0.338901	-0.193323	-0.014221	0.313743	0.227657	-0.132642	-0.284652	-0.104796	-0.200931	0.378753	0
891e2042323ffff	-0.578869	0.324108	-0.507577	-0.754409	-0.084290	0.240572	-0.252477	0.415875	-0.184438	-0.246741	4
891e2051873ffff	-0.310405	0.045352	-0.296344	-0.414064	-0.021333	0.244782	-0.422900	0.357396	-0.103730	0.055133	1
891e2042e2bffff	-0.785091	0.567634	0.190860	-0.089924	-0.219985	0.231418	0.420068	0.015617	0.521316	-0.284058	3

3168 rows × 11 columns