Skip to content

Model of pedestrian flows through New York City

Notifications You must be signed in to change notification settings

ATFutures/nyped

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Build Status Project Status: Concept - Minimal or no implementation has been done yet.

nyped

Model of pedestrian flows against empirical pedestrian counts for New York City, constructed from “flow layers” formed from pair-wise matching between the following seven categories of origins and destinations:

  1. subway
  2. residential
  3. transportation
  4. sustenance
  5. entertainment
  6. education
  7. healthcare

An eighth category is network centrality, with additional layers modelling dispersal from each of these categories. The model explains R2= 83.9 of the observed variation in pedestrian counts. Final results, with significantly explanatory layers named according to the first three letters of the above categories, looks like this:

Layer Name Estimate Std. Error t value Pr(>t)
edu-tra 23977 4484 5.35 0.0000
edu-sus 16904 5572 3.03 0.0031
edu-dis -78057 24521 -3.18 0.0020
edu-hea -24921 4445 -5.61 0.0000
ent-tra 38179 12019 3.18 0.0020
hea-dis 105658 10706 9.87 0.0000
sub-dis 23 3 8.99 0.0000
sub-hea 8 1 6.66 0.0000
sub-tra 6 1 5.08 0.0000
sub-cen -10 1 -6.99 0.0000
sus-res 6258 1232 5.08 0.0000
sus-ent 1446 361 4.00 0.0001
sus-sub -1337 331 -4.04 0.0001
sus-edu -5924 978 -6.06 0.0000

Table 1. Statistical parameters of final model of pedestrian flows through New York City.

A sample of actual flows looks like this:

And a final statistical relationship between modelled and observed pedestrian counts looks like this:

About

Model of pedestrian flows through New York City

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages