 |
 |
Dear all, I've been looking to find Space Syntax's normalisations to account for different sized graphs/sub-graphs for different impedances (angular, metric and topological) and measures (choice/betweenness and integration/closeness). So far I've found: Topological integration: 1/(TD / NC^2) Angular Choice: Log(Choice+1) / Log(TD+3) Angular Integration: (NC^1.2)/TD [TD = Total_Depth, NC=Node_Count] Are there equivalents for Topological Choice (I know it was not widely used at that time because Integration was more successful); and Choice and Integration using metric distance? (I did see a message on here from Bill Hillier (Subject: "Mean depth in Depthmap?" on 15/08/2011 where he suggested topological normalisation works for angular measures - perhaps this can be used? Also, is the above formula for Topological Integration the one which is widely/should be used? I know the formula involving D-values produces very similar results. Thank you in advance and all the best, Michael