An R Interface to the GraphHopper Directions API

The purpose of {graphhopper} is to provide a quick and easy access to the GraphHopper Directions API. Responses can be converted into simple feature (sf) objects in a convenient way. The package is not a complete wrapper of the API. Currently it mainly supports the API also included in Graphhopper’s Open Source routing engine. New features will be added continuously.

Dive into the documentation here.

Installation

Install the release version from CRAN with:

install.packages("graphhopper")

Install the development version from GitHub with:

# install.packages("remotes")
remotes::install_github("crazycapivara/graphhopper-r")

Get started

Run your own GraphHopper instance (with data of Berlin):

Setup

library(graphhopper)

API_URL <- "http://localhost:8989"
gh_set_api_url(API_URL)

info <- gh_get_info()

info$version
#> [1] "2.0"
info$data_date
#> [1] "2020-12-04T21:42:03Z"
gh_bbox(info)
#> [1] 13.06979 52.33306 13.76352 52.67962

Route

Get a route in Berlin:

start_point <- c(52.592204, 13.414307)
end_point <- c(52.539614, 13.364868)

(route <- gh_get_route(list(start_point, end_point)) %>%
    gh_as_sf())
#> Simple feature collection with 1 feature and 2 fields
#> geometry type:  LINESTRING
#> dimension:      XY
#> bbox:           xmin: 13.36501 ymin: 52.53949 xmax: 13.41483 ymax: 52.59234
#> CRS:            EPSG:4326
#>     time distance                       geometry
#> 1 697420 7541.438 LINESTRING (13.41422 52.592...

ggplot(data = route) +
  geom_sf() +
  theme(axis.text.x = element_text(angle = 45))


route$time
#> [1] 697420

via_point <- c(52.545461, 13.435249)

route2 <- gh_get_route(list(start_point, via_point, end_point))

gh_time_distance(route2)
#> $time
#> [1] 1168950
#> 
#> $distance
#> [1] 12843.67

ggplot(data = gh_as_sf(route2)) +
  geom_sf() +
  theme(axis.text.x = element_text(angle = 45))


gh_points(route2) %>%
  head()
#>        lon      lat gh_id
#> 1 13.41422 52.59234     0
#> 2 13.41321 52.59212     1
#> 3 13.41483 52.58964     2
#> 4 13.41539 52.59004     3
#> 5 13.41599 52.59032     4
#> 6 13.41942 52.59145     5

gh_instructions(route2)[, c("lon", "lat", "gh_id", "gh_end_id", "text", "distance")] %>%
  head()
#>        lon      lat gh_id gh_end_id                                   text
#> 1 13.41422 52.59234     0         1        Continue onto Buchholzer Straße
#> 2 13.41321 52.59212     1         2       Turn left onto Buchholzer Straße
#> 3 13.41483 52.58964     2         5 Turn sharp left onto Buchholzer Straße
#> 4 13.41942 52.59145     5         8         Turn right onto Grumbkowstraße
#> 5 13.42352 52.58588     8        15    Turn left onto Blankenburger Straße
#> 6 13.43019 52.58851    15        18      Turn right onto Pasewalker Straße
#>   distance
#> 1   72.248
#> 2  296.761
#> 3  373.025
#> 4  678.120
#> 5  556.120
#> 6  619.849

Shortest path tree

start_point <- c(52.53961, 13.36487)

points_sf <- gh_get_spt(start_point, time_limit = 180) %>%
  gh_as_sf() %>%
  dplyr::mutate(time = (time / 1000 / 60))

ggplot() +
  geom_sf(data = points_sf, aes(colour = time), size = 0.5) +
  theme(axis.text.x = element_text(angle = 45))

Also query previous nodes to plot the network:

(columns <- gh_spt_columns(
  prev_longitude = TRUE,
  prev_latitude = TRUE,
  prev_time = TRUE
))
#> [1] "longitude"      "latitude"       "time"           "distance"      
#> [5] "prev_longitude" "prev_latitude"  "prev_time"

lines_sf <- gh_get_spt(end_point, time_limit = 240, columns = columns) %>%
  dplyr::mutate(mean_time = ((time + prev_time) / 2) / 1000 / 60) %>%
  gh_spt_as_linestrings_sf()

ggplot() +
  geom_sf(data = lines_sf, aes(color = mean_time), size = 1) +
  theme(axis.text.x = element_text(angle = 45))

Isochrone

start_point <- c(52.53961, 13.36487)

isochrone_sf <- gh_get_isochrone(start_point, time_limit = 180) %>%
  gh_as_sf()

ggplot() +
  geom_sf(data = isochrone_sf, fill = "yellow") +
  geom_sf(data = points_sf, aes(colour = time), size = 0.5) +
  theme(axis.text.x = element_text(angle = 45))