https://github.com/mpadge/bike-clusters

:bike: Analyses of data from public hire bicycles
https://github.com/mpadge/bike-clusters

Last synced: about 2 months ago
JSON representation

:bike: Analyses of data from public hire bicycles

• Host: GitHub
• URL: https://github.com/mpadge/bike-clusters
• Owner: mpadge
• License: gpl-2.0
• Created: 2014-08-12T20:45:27.000Z (almost 11 years ago)
• Default Branch: master
• Last Pushed: 2015-12-18T10:25:35.000Z (over 9 years ago)
• Last Synced: 2025-02-14T13:24:45.088Z (4 months ago)
• Language: C++
• Homepage:
• Size: 399 KB
• Stars: 2
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# BikeClusters

A collection of programs to analyse clusters within urban areas as produced by

usage patterns of hire bicycle systems. Requires the output of

[`bike-correlations`](https://github.com/mpadge/bike-correlations), in

particular the correlation (R2) and distance matrices.  Currently able to

analyse data from London, NYC, Boston, Chicago, and Washington DC, using cluster

methods of `ward`, `complete`, `k-means`, and `skater`.

Analyses are ultimately based on comparisons of the total distances ridden

within clusters to equivalent distances ridden between them.

Use makefile to build. Note that the C++ routines can take a long time to

execute, and so are built as stand-alones, rather than being integrated into R.

The two routines are `ClustersNeutral`, which generates neutrally expected

values of inter- and intra-cluster distance, and `Clusters Actual`, which does

the corresponding calculations for the observed rides.

All calculations are repeated for a range of numbers of clusters (up to 100,

except skater only up to 50).

### Usage for a given city and clustering method:

1. `>./ClustersNeutral city`

2. `R> get.clusters (city method)`

3. `R> get.skater.groups (city, method)`

4. `>./ClustersActual city`

5. `R> calc.pnc (city, method`

6. `R> clust.sig (city, method)`

7. `>/randomClusters` city direction (0/1)

See `aaaread-this` for further details.

### build:

1. cd ./build  

2. cmake ..  

3. make

[![Build

Status](https://travis-ci.org/mpadge/bike-clusters.svg?branch=master)](https://travis-ci.org/mpadge/bike-clusters)

Build fails because of CGAL/Exact_predicates!