Street Network Alignment Analysis for Origin-Destination Pairs Across 100 Cities

77,293 origin-destination pairs from 100 cities are analyzed to evaluate a novel Earth Mover's Distance method for measuring alignment with street networks. The dataset supports research on predicting route length and complexity based on orientation distributions. It was authored by Arvid Krantz-Horned and last updated in April 2026.

Use Cases

• Predict route length from origin-destination alignment scores using Earth Mover's Distance.
• Analyze correlation between street network grid-likeness and the predictive power of alignment heuristics.
• Compare orientation distributions across 100 cities to classify street network typologies.
• Validate the hypothesis that misalignment correlates with more complex routes for specific city subsets.

Strengths

• Contains 77,293 origin-destination pairs for statistical analysis.
• Covers 100 cities with diverse street network structures for comparative study.
• Dataset size of 496.4 MB indicates detailed feature representation.

Limitations

• Specific column names and row count per city are not provided.
• Potential geographic bias depending on the selection of the 100 cities.
• Methodology relies on a specific Earth Mover's Distance implementation which may limit direct comparability.

Provenance

Source

Figshare, authored by Arvid Krantz-Horned.

Collection Method

Created for research evaluating a novel alignment metric using Earth Mover's Distance on orientation distributions.

Freshness

Last updated in April 2026.

Geography

100 unspecified cities.