Example notebooks

Three tiers of runnable example notebooks ship with aperta. Each is paired with a .py source file via Jupytext in the repository; the rendered versions below are the committed .ipynb outputs.

Tier	Notebook	Run time	What it covers
Minimal	minimal	~10 s	Every aperta primitive exercised exactly once, OSM data only (Cambridge MA). The “what does aperta do?” demo.
Walkthrough	walkthrough	~1 min	Guided tour of every primitive on real OSM data (Central Paris): tiered ODs, geo-keyed reindex, overheads, three accessibility metrics, path-first per-edge feature aggregation, cross-modal logsum. Covers the full accessibility surface end-to-end.
Calibration	Calibration — traffic flows + edge weights	~30 min	Consolidated calibration workflow on the Canton of Zurich (fetched inline from OSM): traffic-flow estimation via nested-betweenness sampling with survey-derived P(C), then flow-aware edge-weight calibration against Google-Maps peak-hour travel times.

The calibration notebook fetches the OSM car network for the Canton of Zurich inline via osmnx; the only external inputs are ground-truth files whose source terms preclude redistribution — Google-Maps-derived car travel times and Swiss ASTRA counter readings. It renders here from committed outputs; re-running locally requires the private ground-truth inputs. A public-data version is planned.