Examples¶

Visual showcase¶

The cover figure below is generated by the examples/about-the-data/ example, which exercises every packaged boundary layer and spatial helper in a single script.

nyc-geo-toolkit cover figure

Panel breakdown¶

Panel	Layer	Features	API demonstrated
Boroughs	`borough`	5	`load_nyc_boundaries_geodataframe()`, `to_web_mercator()`, `add_osm_basemap()`
Community Districts	`community_district`	71	`load_nyc_boundaries_geodataframe()`
Council Districts	`council_district`	51	`load_nyc_boundaries_geodataframe()`
NTAs	`neighborhood_tabulation_area`	195	`load_nyc_boundaries_geodataframe()`
Census Tracts (clipped)	`census_tract`	varies	`load_nyc_boundaries()`, `clip_boundaries_to_bbox()`
Geodesy Helpers	n/a	3 circles	`build_circle_polygon()`, `walk_radius_meters()`, `haversine_distance_meters()`

Generating the figure¶

cd examples/about-the-data
uv run python main.py
# -> artifacts/nyc-geo-toolkit-cover.png

Self-contained examples¶

Each example in the examples/ directory is a standalone project with its own pyproject.toml and can be run independently.

About the Data¶

Kitchen-sink visualization of every boundary layer and spatial helper. Generates a multi-panel cover figure.

cd examples/about-the-data && uv run python main.py

Boundary Quickstart¶

Loads a single borough boundary and exports it as GeoJSON.

cd examples/boundary-quickstart && uv run python main.py

Normalization Demo¶

Shows how nyc-geo-toolkit normalizes inconsistent layer names and geography values into canonical forms.

cd examples/normalization-demo && uv run python main.py

Boundary Explorer Tearsheet¶

Interop showcase for the random-walks ecosystem. Loads community-district boundaries via load_nyc_boundaries("community_district"), builds a synthetic factor_factory.tidy.Panel with a TreatmentEvent, fits a two-way fixed-effects DiD via factor_factory.engines.did.estimate(..., methods=("twfe",)), and renders a factor_factory.jellycell.tearsheets.findings manuscript.

Pulls in factor-factory and jellycell — they are NOT default nyc-geo-toolkit dependencies, only extras for this example.

cd examples/boundary-explorer-tearsheet && uv sync && uv run python main.py

The synthetic outcome has a known 4.0-unit additive treatment effect, so the fitted ATT should land near 4.0 — making this both a pattern demo and a smoke test. Downstream packages (nyc311, subway-access) can follow the same template with their real outcomes.