Visualizing Street Orientations of Cities Worldwide Using Python and OSMnx

Overview

The article explores the use of Python and the OSMnx library to analyze and visualize the street network orientations of cities around the world. It presents a series of polar histograms, also known as "rose plots," that display the distributions of street angles for various cities, revealing their underlying grid patterns and urban spatial order.

Method

The author uses the OSMnx library, a Python package for working with OpenStreetMap data, to download and model the street networks of different cities.
The street orientations are calculated and visualized as polar histograms, where the angle represents the orientation of the streets, and the length of the bars represents the relative frequency of streets at that angle.
The visualizations reveal the distinctive grid patterns of cities, ranging from highly ordered grids (e.g., Toronto, Barcelona) to more chaotic or irregular patterns (e.g., Seoul, Mumbai).

Examples

The article includes several examples of city street orientation visualizations, including:

25 cities from around the world, such as Bangkok, Barcelona, Beijing, Cairo, Delhi, Dubai, Glasgow, Hong Kong, Lagos, London, Madrid, Melbourne, Mexico City, Moscow, Mumbai, Munich, Paris, Rio de Janeiro, Rome, Seoul, Sydney, Tehran, Toronto, Warsaw, and Tokyo.
26 cities from the United States, including Atlanta, Boston, Buffalo, Charlotte, Chicago, Cleveland, Dallas, Denver, Detroit, Houston, Las Vegas, Los Angeles, Manhattan, Miami, Minneapolis, Orlando, Philadelphia, Phoenix, Portland, Sacramento, San Francisco, Seattle, St. Louis, Tampa, and Washington D.C.

Further Analysis

The article also mentions a research paper by the author that analyzes and visualizes the street network orientations of 100 cities around the world, discussing the methods, entropy calculations, and images in more detail. The paper can be accessed at https://geoffboeing.com/publications/urban-spatial-order-entropy/.

Comments

The article includes numerous comments from readers, expressing interest in seeing visualizations for additional cities, such as Tokyo, Kyoto, Chandigarh, New Orleans, Amsterdam, Bogotá, and Singapore, among others. The author provides links to the full analysis of 100 cities, which includes many of the requested cities.

Other comments discuss potential factors influencing street orientations, such as geographic features (coastlines, rivers), historical planning and development patterns, and the influence of magnetic north in early surveying and city planning.

Code

The article includes a link to the GitHub repository containing the Jupyter Notebook used to generate the street orientation visualizations: https://github.com/gboeing/osmnx-examples/blob/master/notebooks/17-street-network-orientations.ipynb

# Example code snippet to generate street orientation visualization
import osmnx as ox
import matplotlib.pyplot as plt

# Download street network data for a city
place = 'Barcelona, Catalonia, Spain'
graph = ox.graph_from_place(place, network_type='drive')

# Calculate street orientations
edges = ox.graph_to_gdfs(graph, edges=True)
edges['angle'] = edges.apply(ox.bearing, geom1='u', geom2='v', auxiliary_geometries='edges')

# Visualize street orientations as a polar histogram
fig, ax = plt.subplots(figsize=(10, 10), subplot_kw={'projection': 'polar'})
edges['angle'].hist(bins=36, density=True, ax=ax, alpha=0.7)
ax.set_theta_zero_location('N')
ax.set_thetalim(0, 360)
ax.set_title(f'Street network orientation in {place}')
plt.show()

The code demonstrates how to download street network data for a specified city using OSMnx, calculate the street orientations (angles), and visualize them as a polar histogram using Matplotlib.

Summarized Content