usr110/pct

Development of a 'Propensity to Cycle Tool'

<script type="text/javascript" src="http://www.math.union.edu/~dpvc/transfer/mathjax/mathjax-in-github.user.js"></script>

pct: Propensity to cycle tool

This repository allows users to estimate the 'propensity to cycle' between different origin-destination pairs.

The project is funded by the Department for Transport (DfT) so the initial case studies will be taken from the UK. However, it is expected that the methods will be of use elsewhere. For that reason, attempts have been made to make the examples generalisable. All examples presented here are reproducible using code in this repository and data stored in the pct-data repository.

So, if you run the following lines of code on your computer from within this folder, you should get the same result. Reproducible research!

A simple example

# system("git clone [email protected]:Robinlovelace/pct-data.git") # see set-up.R
source("set-up.R")
# load some flow data
fleeds <- read.csv("pct-data/leeds/sample-leeds-centre-dists.csv")
# load the zones
leeds <- readOGR("pct-data/leeds/", "leeds-central-sample")

## OGR data source with driver: ESRI Shapefile 
## Source: "pct-data/leeds/", layer: "leeds-central-sample"
## with 25 features and 3 fields
## Feature type: wkbPolygon with 2 dimensions

Now we can estimate propensity to cycle, by using the distance decay function from (Iacono et al. 2010):

where $\alpha$, the proportion of made for the shortest distances and $\beta$, the rate of decay are parameters to be calculated from empirical evidence.

To implement this understanding in R code we can use the following function:

# Distance-dependent mode switch probs
iac <- function(x, a = 0.3, b = 0.2){
  a * exp(1)^(-b * x)
}

Apply this function to openly accessible flow data:

fleeds$p_cycle <- iac(fleeds$dist / 1000)
fleeds$n_cycle <- fleeds$p_cycle * fleeds$All.categories..Method.of.travel.to.work
fleeds$pc1 <- fleeds$n_cycle - fleeds$Bicycle

Now we can create a simple visualisation of the result:

plot(leeds)

for(i in which(fleeds$Area.of.residence == leeds$geo_code[1])){
  from <- leeds$geo_code %in% fleeds$Area.of.residence[i]
  to <- leeds$geo_code %in% fleeds$Area.of.workplace[i]
  x <- coordinates(leeds[from, ])
  y <- coordinates(leeds[to, ])
  lines(c(x[1], y[1]), c(x[2], y[2]), lwd = fleeds$pc1[i] )
}