• Background
• Using stplanr
• Future plans

• Contracted to create the Propensity to Cycle Tool (PCT)
• Best way to understand stplanr is via a live demo, at www.pct.bike
• Needed functions to work with origin-destination data and routing services

Common computational tasks in transportation planning:

• Access to and processing of data on transport infrastructure and behaviour
• Analysis and visualisation of origin-destination flow data
• Allocation of flows to the transport (road) network
• Development of models to estimate transport behaviour
• The calculation of 'catchment areas' affected by transport infrastructure

Transport planning is notoriously reliant on 'black boxes' and the same applies to scientific research into transport systems [@Waddell2002].

stplanr seeks to address these issues.

stplanr lives here: https://github.com/ropensci/stplanr

Package can be installed from CRAN or GitHub (see the package's README for details), it can be loaded in with library():

install.packages("stplanr") # stable CRAN version
# devtools::install_github("ropensci/stplanr") # dev version

library(stplanr) # also loads spatial package

## Loading required package: sp

• Dev version requires rtools on Windows

• Assisted with documentation: ropensci/stplanr/commits?author=karthik
• Set-up Continuous Integration (CI): https://travis-ci.org/ropensci/stplanr

• Provided peer review of the code: https://github.com/ropensci/onboarding/issues/10

Use @importFrom whenever possible. Right now you have import(openxlsx) and import(sp) in your NAMESPACE file. Just import the functions you need. Same for other pkg deps.

Tests: Pleae write tests to cover at least all the major functions before we accept. Use testthat::skip_on_cran() for any tests that do web requests, so that CRAN tests don't fail in case a service is temporarily down

I think token's can be a bit easier for the user. Right now you have e.g.,

if (!Sys.getenv('CYCLESTREET') == "") {
    cckey <- Sys.getenv('CYCLESTREET')
}
if(is.null(cckey)){
    stop("You must have a CycleStreets.net api key saved as 'cckey'")
}

data("flow", package = "stplanr")
head(flow[c(1:3, 12)])

##        Area.of.residence Area.of.workplace All Bicycle
## 920573         E02002361         E02002361 109       2
## 920575         E02002361         E02002363  38       0
## 920578         E02002361         E02002367  10       0
## 920582         E02002361         E02002371  44       3
## 920587         E02002361         E02002377  34       0
## 920591         E02002361         E02002382   7       0

data("cents", package = "stplanr")
as.data.frame(cents[1:3,-c(3,4)])

##       geo_code  MSOA11NM coords.x1 coords.x2
## 1708 E02002384 Leeds 055 -1.546463  53.80952
## 1712 E02002382 Leeds 053 -1.511861  53.81161
## 1805 E02002393 Leeds 064 -1.524205  53.80410

flow_single_line = flow[4,] # select only the first line
desire_line_single = od2line(flow = flow_single_line, zones = cents)
plot(cents)
plot(desire_line_single, add = T)

• We selected a single 'od-pair' (flow[4,])
• The function od2line() matched the cents matching the lines and created a line (the hard bit)
• How? Check the source code!

od2line

## function (flow, zones, destinations = NULL, zone_code = names(zones)[1], 
##     origin_code = names(flow)[1], dest_code = names(flow)[2], 
##     zone_code_d = NA, silent = TRUE) 
## {
##     l <- vector("list", nrow(flow))
##     if (is.null(destinations)) {
##         if (!silent) {
##             message(paste("Matching", zone_code, "in the zones to", 
##                 origin_code, "and", dest_code, "for origins and destinations respectively"))
##         }
##         for (i in 1:nrow(flow)) {
##             from <- zones@data[[zone_code]] %in% flow[[origin_code]][i]
##             if (sum(from) == 0) 
##                 warning(paste0("No match for line ", i))
##             to <- zones@data[[zone_code]] %in% flow[[dest_code]][i]
##             if (sum(to) == 0 & sum(from) == 1) 
##                 warning(paste0("No match for line ", i))
##             x <- sp::coordinates(zones[from, ])
##             y <- sp::coordinates(zones[to, ])
##             l[[i]] <- sp::Lines(list(sp::Line(rbind(x, y))), 
##                 as.character(i))
##         }
##     }
##     else {
##         if (is.na(zone_code_d)) {
##             zone_code_d <- names(destinations)[1]
##         }
##         if (!silent) {
##             message(paste("Matching", zone_code, "in the zones and", 
##                 zone_code_d, "in the destinations,\nto", origin_code, 
##                 "and", dest_code, "for origins and destinations respectively"))
##         }
##         for (i in 1:nrow(flow)) {
##             from <- zones@data[[zone_code]] %in% flow[[origin_code]][i]
##             if (sum(from) == 0) 
##                 warning(paste0("No match for line ", i))
##             to <- destinations@data[[zone_code_d]] %in% flow[[dest_code]][i]
##             if (sum(to) == 0 & sum(from) == 1) 
##                 warning(paste0("No match for line ", i))
##             x <- sp::coordinates(zones[from, ])
##             y <- sp::coordinates(destinations[to, ])
##             l[[i]] <- sp::Lines(list(sp::Line(rbind(x, y))), 
##                 as.character(i))
##         }
##     }
##     l <- sp::SpatialLines(l)
##     l <- sp::SpatialLinesDataFrame(l, data = flow, match.ID = FALSE)
##     sp::proj4string(l) <- sp::proj4string(zones)
##     l
## }
## <environment: namespace:stplanr>

o = cents[cents$geo_code %in% flow$Area.of.residence[4],]
d = cents[cents$geo_code %in% flow$Area.of.workplace[4],]
l_single = Lines(list(Line(rbind(o@coords, d@coords))), ID = 1)
l_sp = SpatialLines(LinesList = list(l_single))

plot(cents)
points(o, cex = 5)
points(d, cex = 5)
flow[4, 1:3]

##        Area.of.residence Area.of.workplace All
## 920582         E02002361         E02002371  44

plot(l_sp, add = TRUE)

l <- od2line(flow = flow, zones = cents)
# remove lines with no length
l <- l[!l$Area.of.residence == l$Area.of.workplace,]
plot(l, lwd = l$All / 10)

stplanr has various functions for route allocation:

route_cyclestreet() # UK, cycling
route_graphhopper() # worldwide, any mode
route_transportapi_public() # UK, public transport
viaroute() # worldwide, any mode

library(leaflet)
leaflet() %>% addTiles() %>% addPolylines(data = r)

routes_fast = line2route(l = l)

plot(l)
plot(routes_fast, add = T, col = "red")

• Add more services (e.g. OpenTripPlanner)
• Integrate interface
• Add support for OSM data download and analysis
• Via interface to osmdatar

• Work with practitioners to make it more useful for them (e.g. ITP)
• Link with industry (e.g. ITP)
• Make more international
• A global propensity to cycle tool?
• Better handling of GPS data

• Download and test the dev version with:

devtools::install_github("ropensci/stplanr")

• Create issues on our issue tracker
• Check out the vignette
• See my tutorial on visualising spatial data for eRum: Creating-maps-in-R/blob/master/vignettes/vspd-base-shiny.md
• Academic paper on the PCT: http://arxiv.org/abs/1509.04425

• Work through the vignette provided by:

vignette("introducing-stplanr")

• Create a single desire line for the 12th flow without using stplanr code (stplanr solution below):

od2line(flow = flow[12,], zones = cents)

• Visualise the flowlines object using tmap with different colours and widths
• Create a 'oneway' version of the flowlines and visualise these (tricky)
• Create a sf version of the flowlines object (advanced)