Methods for calculating cognitive distance and accessibility from global GIS data

The main methods are found in the Python Jupyter notebook ('Cognitive Distance Anon'). This includes methods for estimating the effect of urban features (landmarks, land uses), intersections, turns, and network density on cognitive distance. The notebook clearly highlights the parameters used in defining the effect of each facet. The notebook also contains methods for calculating cognitive distance for a range of cities, drawing down GIS data for each city from OpenStreetMap using the OSMNx library. For each city, 500 random routes are calculated and distances extracted. The notebook contains additional methods relating to the generation of data visualisations used in the paper.

The calculation of cognitive distance is supported by an additional functional package, landmark_functions.py. This code contains additional methods for landmark identification from OpenStreetMap data. This classification is based on the physical, pragmatic, and cultural components of each building within the dataset.

The text file (test_cities.txt) contains the cities and coordinates used in estimation of cognitive distance, as documented in the paper.