A typology of street patterns

摘要: We propose a quantitative method to classify cities according their street pattern. use the conditional probability distribution of shape factor blocks with given area and define what could constitute ‘fingerprint’ city. Using simple hierarchical clustering method, these fingerprints can then serve as basis for typology cities. apply this set 131 in world, at an intermediate level dendrogram, we observe four large families characterized by different abundances certain shape. At lower classification, find that most European American our sample fall own sub-category, highlighting quantitatively differences between typical layouts both regions. also show example New York its boroughs, fingerprint city be seen sum ones characterizing neighbourhoods inside This provides comparison urban patterns, which helpful better understanding causes mechanisms behind distinct shapes.