API reference¶

Aperta is organized as a flat collection of algorithm modules. Each module focuses on one phase of the six-phase accessibility workflow and operates on plain numpy / pandas / networkx inputs without any framework scaffold.

At a glance¶

Module	Purpose
`od_pairs`	Tiered OD pair structures + builders, including cross-modal alignment.
`routing`	Shortest paths on `networkx` graphs via `scipy.sparse.csgraph.dijkstra`: tiered OD routing, per-edge / per-node feature aggregation along realized paths, intrazonal-cost flooring.
`accessibility`	Cumulative-opportunity, gravity, nearest-k metrics. Per-node or per-cell output depending on input ODM class.
`utility`	Linear utility specifications and the routing-+-endpoint pipeline for utility-based costs; consumed by `gravity` with an exp decay for logsum accessibility.
`overhead`	First/last-mile overheads on cost ODMs (node-keyed and geo-keyed), with helpers for zone-tier last-mile aggregation.
`traffic_flows`	Traffic-volume estimation via cost-weighted nested-node sampling.
`calibration`	OLS calibration of per-edge weights against observed travel times; bearing-aware traffic-counter snapping + counter-fit evaluation for flow calibration.
`network_processing`	Network helpers — intersection consolidation, sampled edge betweenness, node snapping, per-direction lane counts.
`geo_processing`	Geometry helpers — H3 grids, line bearings, KDTree-based buffer aggregations, raster sampling.
`geo_mapping`	Spatial-join wrappers between points, polygons, and filtered lines.
`osm_helpers`	OSM data fetching + per-edge categorization via `osmnx`. Requires `aperta[osm]`.
`topography`	Copernicus GLO-30 DEM download + raster sampling. Requires `aperta[topo]`.
`visualization`	Plot helpers — choropleth panels, multi-panel comparisons, per-edge `LineCollection` rendering, styled colorbars.
`errors`	aperta-specific exception types.

Core OD + routing

Accessibility metrics

Traffic flows + calibration

Geographic + network helpers

Misc