1. Introduction

Many transport planning and modeling tasks need travel time estimates between origins and destinations. R5 computes realistic door-to-door travel times in multimodal networks very fast, and r5r offers three functions for it:

• travel_time_matrix()
• expanded_travel_time_matrix()
• arrival_travel_time_matrix()

This vignette shows, with a reproducible example, how they work and how they differ.

2. Build routable transport network with build_network()

We use the Porto Alegre (Brazil) sample data included in r5r.

# increase Java memory
options(java.parameters = "-Xmx2G")

# load libraries
library(r5r)
library(data.table)
library(ggplot2)

# build a routable transport network with r5r
data_path <- system.file("extdata/poa", package = "r5r")
r5r_network <- build_network(data_path)

# routing inputs
mode <- c('walk', 'transit')
max_trip_duration <- 60 # minutes

# departure time
departure_datetime <- as.POSIXct("13-05-2019 14:00:00", 
                                 format = "%d-%m-%Y %H:%M:%S")

# load origin/destination points
points <- fread(file.path(data_path, "poa_points_of_interest.csv"))

3. The travel_time_matrix() function

travel_time_matrix() quickly computes travel times between all origin/destination pairs for a departure time and transport mode. Other parameters include:

• max_trip_duration: maximum trip duration
• max_rides: maximum number of public transport rides
• max_walk_time and max_bike_time: maximum walking or cycling time to and from public transport
• walk_speed and bike_speed: average walking or cycling speed (km/h)
• max_fare: maximum monetary cost in public transport. See this vignette.

# estimate travel time matrix
ttm <- travel_time_matrix(
  r5r_network,   
  origins = points,
  destinations = points,    
  mode = mode,
  max_trip_duration = max_trip_duration,
  departure_datetime = departure_datetime
  )

head(ttm, n = 10)

Travel times can vary significantly across the day with public transport service levels. The time_window and percentiles parameters handle this efficiently: R⁵ computes travel times for departures every minute within time_window and returns the selected percentiles (time window vignette).

4. The expanded_travel_time_matrix() function

expanded_travel_time_matrix() returns more than the total travel time: by default, it also lists the public transport routes taken between each origin/destination pair. With breakdown = TRUE, it adds each trip’s number of public transport rides and its access, waiting, in-vehicle, transfer and egress times, which can be slower for large data sets.

A general call to expanded_travel_time_matrix()

ettm <- expanded_travel_time_matrix(
  r5r_network,   
  origins = points,
  destinations = points,    
  mode = mode,
  max_trip_duration = max_trip_duration,
  departure_datetime = departure_datetime
  )

head(ettm, n = 10)

Calling expanded_travel_time_matrix() with breakdown = TRUE

ettm2 <- expanded_travel_time_matrix(
  r5r_network,   
  origins = points,
  destinations = points,    
  mode = mode,
  max_trip_duration = max_trip_duration,
  departure_datetime = departure_datetime,
  breakdown = TRUE
  )

head(ettm2, n = 10)

Over its time_window (10 minutes by default), expanded_travel_time_matrix() returns the fastest route departing at each minute. This can be very memory intensive for large data sets and time windows.

ettm_window <- expanded_travel_time_matrix(
  r5r_network,   
  origins = points,
  destinations = points,    
  mode = mode,
  max_trip_duration = max_trip_duration,
  departure_datetime = departure_datetime,
  breakdown = TRUE,
  time_window = 10
  )

ettm_window[15:25,]

5. The arrival_travel_time_matrix() function

travel_time_matrix() and expanded_travel_time_matrix() take a departure time. To route by arrival time, use arrival_travel_time_matrix(): given the latest acceptable arrival time and a maximum trip duration, it returns the travel time of the trip with the latest departure that still arrives in time.

This models trips where arriving by a set time matters, such as getting to work or school by 9 a.m.: people usually take the latest departure that gets them there on time, not the fastest trip that leaves them waiting at the destination.

As in expanded_travel_time_matrix(), the output has additional columns (and breakdown = TRUE is available).

A general call to arrival_travel_time_matrix()

arrival_datetime <- as.POSIXct(
 "13-05-2019 14:00:00",
 format = "%d-%m-%Y %H:%M:%S"
)

arrival_ttm <- arrival_travel_time_matrix(
  r5r_network,
  origins = points,
  destinations = points,
  mode = c("WALK", "TRANSIT"),
  arrival_datetime = arrival_datetime,
  max_trip_duration = 60
)

head(arrival_ttm, n = 10)

r5r::stop_r5(r5r_network)
rJava::.jgc(R.gc = TRUE)