Network-based accessibility analysis for X-minute cities, proximity planning, and urban opportunity mapping.

accessX is a Python library for studying what people can reach, how easily they can reach it, and how accessibility is distributed across places and populations.

The X can represent any network cost threshold: walking minutes, cycling time, distance, generalized cost, or a custom edge-weight measure.

Built on GeoPandas, OSMnx, NetworkX, H3, and rasterio, accessX provides a focused API for reproducible accessibility workflows without hiding the underlying geospatial data.

https://app.transformtransport.org/readthedocs/accessx.html

• Plug-n-play, without locking you in: accessX provides ready-to-use methods for common accessibility workflows while keeping the underlying GeoDataFrames and OSMnx graphs available for customization.
• Use the data you have: work with OSM POIs and WorldPop data, or provide your own destinations, capacities, population grids, demographic groups, and demand measures.
• Define accessibility through the cost that matters: estimate walking accessibility with a standard average speed, use slope-sensitive travel time, or assign your own edge cost to an OSMnx graph for factors such as comfort, safety, effort, distance, or a generalized impedance.
• Network-based by default: accessibility is calculated over the street graph rather than straight-line distance.
• Transparent and composable: modules can be used independently, outputs remain reusable GeoDataFrames, and intermediate layers can be saved, inspected, and replaced.
• Built for reproducible long-running analysis: retries, progress bars, clear errors, and graph/data persistence support workflows that can be rerun and audited.

• How many services are reachable within 5, 10, or 15 minutes?
• What is the network travel cost to the nearest healthcare facility or park?
• How does accessibility change when nearby opportunities are weighted more strongly?
• Where is service supply high or low relative to population demand?
• How many people can access the same POI?
• How evenly is accessibility distributed across neighborhoods or population groups?

Area of interest
    -> H3 hex grid
    -> walking or cycling network
    -> edge travel costs
    -> OSM POIs and population demand
    -> accessibility scores
    -> co-accessibility and equity analysis

Isochrones are available as an optional communication and visualization layer.

The case-study notebooks save their intermediate and final GeoDataFrames, so figures can be regenerated from existing outputs without rerunning network routing or data downloads.

Population demand is allocated to H3 hexagons and used as the demand surface for supply-demand accessibility models such as 2SFCA.

OpenStreetMap features are grouped into analytical opportunity categories while preserving their original OSM tags and identifiers.

Equity workflows can summarize whether each origin satisfies a basket of minimum accessibility thresholds.

Co-accessibility identifies parks and public squares where children, adults, and older adults have walkable access, using Shannon diversity to highlight stronger potential intergenerational encounter opportunities.

pip install accessx

Python 3.10 or newer is required.

For local development:

pip install -e .

The library documentation is a static HTML page at docs/access.html.

Open it locally:

open docs/access.html

For FTP hosting, upload docs/access.html, docs/assets/, and any referenced files under docs/figures/.

import accessx as acx
import osmnx as ox

# 1. Area of interest and origins
aoi = ox.geocode_to_gdf("Municipality of Athens, Greece")
hexes = acx.make_hex_grid(aoi, resolution=9)

# 2. Walking network and travel-time cost
graph = acx.build_network(
    aoi,
    city_epsg=2100,
    buffer_m=1000,
    network_type="walk",
    show_progress=True,
)
graph = acx.add_time_cost_constant_speed(
    graph,
    speed_kmh=4.5,
    cost_col="walk_time",
)

# 3. POIs with custom analytical categories backed by explicit OSM tags
pois = acx.get_pois_osm(
    aoi,
    poi_groups={
        "healthcare": {
            "amenity": ["pharmacy", "clinic", "doctors"],
        },
        "open_space": {
            "leisure": ["park", "playground", "garden"],
        },
    },
    show_progress=True,
)

# 4. Cumulative opportunities within 15 walking minutes
counts = acx.count_accessible_pois(
    graph,
    hexes.to_crs(2100),
    pois.to_crs(2100),
    max_cost=15,
    cost_attr="walk_time",
    show_progress=True,
)

The core accessibility functions route over any numeric edge cost on the network. accessX provides built-in helpers for common travel-cost definitions, such as constant-speed travel time and slope-sensitive walking time.

For other interpretations of cost, users can define their own edge-weight function with add_edge_cost. This makes it possible to route over custom measures such as distance, comfort penalties, safety scores, effort, or generalized impedance, as long as they are encoded as numeric edge costs.

For example, you can use a built-in travel-time helper:

graph = acx.add_time_cost_constant_speed(
    graph,
    speed_kmh=4.5,
    cost_col="walk_time",
)

Or define your own cost function:

def comfort_cost(edge):
    return edge["length"] * edge.get("comfort_penalty", 1.0)

graph = acx.add_edge_cost(
    graph,
    cost_fn=comfort_cost,
    cost_col="comfort_cost",
)

The same origins and destinations can then be evaluated through different definitions of accessibility simply by changing cost_attr.

All routing models support custom edge cost attributes and optional progress bars for long calculations.

accessX includes tools for moving from accessibility scores to equity questions: who benefits, where thresholds are met, and how unevenly access is distributed.

A, P, gini, sorted_vals = acx.calculate_lorenz(
    ["count_healthcare", "nearest_cost_healthcare_1"],
    gdf,
    weights="population",
)

sufficient = acx.compute_sufficientarian_score(
    gdf,
    thresholds_ge={
        "count_healthcare": 1,
        "count_daily_needs": 3,
        "count_public_transport": 1,
    },
    thresholds_le={
        "nearest_cost_healthcare_1": 15,
    },
)

Lorenz/Gini workflows summarize inequality across places or population-weighted groups. Sufficientarian scores evaluate whether each origin meets explicit minimum accessibility standards.

POIs and population are treated as transparent inputs to the accessibility models rather than hidden assumptions.

• get_pois_osm queries explicit OSM tags or custom analytical groups, while preserving the original OSM keys, values, feature IDs, and geometries.
• Population workflows can download and clip WorldPop rasters, convert rasters to vector grids, and aggregate one or more population-group columns to H3 hexagons.
• Area-weighted overlap is used for population aggregation, so raster or polygon population cells are split proportionally across intersecting hexes.

Define the travel cost used by network routing.

• add_time_cost_constant_speed: add travel time using a constant speed
• add_slope_based_time: create a Tobler hiking-time cost function
• add_edge_cost: add any custom numeric edge cost

Calculate origin-based and destination-based accessibility.

• count_accessible_pois: cumulative opportunity measure by category
• compute_nearest_poi_cost: nearest POI costs in list, long, or wide output formats
• compute_hansen_accessibility: distance-decayed opportunity accessibility
• compute_2sfca_accessibility: supply-demand catchment accessibility with binary or exponential decay
• compute_co_accessibility: accessible population around each POI, using cumulative or Hansen-style decay

Assess how accessibility is distributed.

• calculate_lorenz: calculate Lorenz curves and Gini indices
• plot_lorenz_curves: visualize accessibility distributions
• compute_sufficientarian_score: score whether explicit accessibility thresholds are met
• plot_sufficientarian_score: visualize score distributions and attainment levels

Prepare analysis areas and spatial origins.

• load_aoi: load an AOI from a vector file or bounding box, optionally buffer and dissolve it
• make_hex_grid: create an H3 hex grid over an AOI

Build and persist OSM street networks.

• build_network: download, clean, and project a walking, cycling, or other OSMnx network
• save_graph: save graph nodes and edges
• load_graph: rebuild a graph from saved node and edge files

Collect and organize destinations from OpenStreetMap.

• get_pois_osm: query explicit OSM tags or custom POI groups, with retries, reports, and progress
• POIQueryError: raised when one or more OSM queries fail after retries

Prepare population demand data.

• infer_country_from_geometry: identify the country intersecting an AOI or hex grid
• get_worldpop_raster: download and optionally clip a WorldPop raster
• raster_to_population_grid: convert raster cells into a vector population grid
• map_population_to_hexes: aggregate raster population to hexes
• map_population_grid_to_hexes: aggregate one or more population columns from a vector grid

Create walksheds and isochrone geometries.

• calculate_isochrones: generate one or more cost-threshold polygons per origin
• Supports edges and hull polygon methods, serial or parallel execution, progress bars, and optional CSV export

Read and write GeoDataFrames.

• read_gdf
• save_gdf

The notebooks provide focused, reproducible examples:

• case_study_isochrones.ipynb: AOIs, hex grids, street networks, costs, and isochrones
• case_study_population.ipynb: WorldPop and population-to-hex workflows
• case_study_pois.ipynb: OSM tags, custom POI groups, retries, and query reports
• case_study_accessibility.ipynb: cumulative opportunities, nearest POIs, Hansen accessibility, and 2SFCA
• case_study_co_accessibility.ipynb: population access around POIs
• case_study_equity.ipynb: territorial and population-based accessibility equity

accessX is released under the MIT License.

The methods in accessX are connected to the following research and tools. Some works used related computational code, while others provide the conceptual or methodological basis for the library. This list includes previous research done by the Transform Transport team.

• Milias, V., & Psyllidis, A. (2022). Measuring spatial age segregation through the lens of co-accessibility to urban activities. Computers, Environment and Urban Systems, 95, 101829. https://doi.org/10.1016/j.compenvurbsys.2022.101829
• Milias, V., Psyllidis, A., & Bozzon, A. (2024). Bridging or separating? Co-accessibility as a measure of potential place-based encounters. Journal of Transport Geography, 121, 104027. https://doi.org/10.1016/j.jtrangeo.2024.104027
• Milias, V. (n.d.). CTwalkMap [Web mapping tool]. https://miliasv.github.io/CTwalkMap/?city=amsterdam

• Milias, V., Tsigdinos, S., Tzouras, P. G., & Kepaptsoglou, K. (2025). Assessing equitable access in X-minute cities through open spatial data. Environment and Planning B: Urban Analytics and City Science. Advance online publication. https://doi.org/10.1177/23998083251398660
• Milias, V. (n.d.). CThood [Web mapping tool]. https://miliasv.github.io/CThood/?city=athens#11.95/38.00354/23.76933

• Hansen, W. G. (1959). How accessibility shapes land use. Journal of the American Institute of Planners, 25(2), 73–76.
• Luo, W., & Wang, F. (2003). Measures of spatial accessibility to health care in a GIS environment: Synthesis and a case study in the Chicago region. Environment and Planning B: Planning and Design, 30(6), 865–884.

• Albashir, A., Messa, F., Presicce, D., Pedrazzoli, A., & Gorrini, A. (2024). 15min City Score Toolkit - Notebook. In Transform Transport Open Source Collection. Zenodo. https://doi.org/10.5281/zenodo.14231427
• Messa, F., Presicce, D., Pedrazzoli, A., Albashir, A., & Gorrini, A. (2026). 15min City Score - Europe Map. In Transform Transport Open Source Collection. Zenodo. https://doi.org/10.5281/zenodo.19610329

• Abdelfattah, L., Deponte, D., & Fossa, G. (2022). The 15-minute city as a hybrid model for Milan. TeMA – Journal of Land Use, Mobility and Environment, 71-86. https://doi.org/10.6093/1970-9870/8653
• Moreno, C., Allam, Z., Chabaud, D., Gall, C., & Pratlong, F. (2021). Introducing the “15-Minute City”: Sustainability, resilience and place identity in future post-pandemic cities. Smart Cities, 4(1), 93–111. https://doi.org/10.3390/smartcities4010006
• Handy, S. (2020). Is accessibility an idea whose time has finally come? Transportation Research Part D: Transport and Environment, 83, 102319. https://doi.org/10.1016/j.trd.2020.102319