Skip to contents

photon

photon is a simple interface and setup manager of the photon OpenStreetMap geocoder. It features unstructured, structured, and reverse geocoding. The package allows requests to the public API but shines at setting up local instances to enable high-performance offline geocoding.

Installation

To install the package from CRAN:

You can install the development version of photon from GitHub with:

# install.packages("remotes")
remotes::install_github("jslth/photon")

Usage

When loading photon, the package assumes that you want send geocoding requests to the public photon API. If you want to change this, you can use the workhorse function new_photon(). Otherwise, you can directly start geocoding.

library(photon)
places <- c("Paris", "Shenzen", "Sao Paulo", "Kinshasa")

cities1 <- geocode(places, layer = "city")
cities1
#> Simple feature collection with 4 features and 12 fields
#> Geometry type: POINT
#> Dimension:     XY
#> Bounding box:  xmin: -46.63338 ymin: -23.55065 xmax: 114.0545 ymax: 48.8535
#> Geodetic CRS:  WGS 84
#> # A tibble: 4 × 13
#>     idx osm_type  osm_id osm_key osm_value type  countrycode name  state country
#>   <int> <chr>      <int> <chr>   <chr>     <chr> <chr>       <chr> <chr> <chr>  
#> 1     1 R          71525 place   city      city  FR          Paris Isla… France 
#> 2     2 R        3464353 place   city      city  CN          Shen… Guan… China  
#> 3     3 R         298285 place   municipa… city  BR          São … São … Brazil 
#> 4     4 R         388103 bounda… administ… city  CD          Kins… Kins… Democr…
#> # ℹ 3 more variables: extent <list>, county <chr>, geometry <POINT [°]>

Structured geocoding means providing address / place details in a structured format.

places_str <- data.frame(
  city = places,
  countrycode = c("FR", "CN", "BR", "CD")
)
cities2 <- structured(places_str)
cities2
#> Simple feature collection with 4 features and 13 fields
#> Geometry type: POINT
#> Dimension:     XY
#> Bounding box:  xmin: -35.76483 ymin: -5.894168 xmax: 114.0545 ymax: 46.91503
#> Geodetic CRS:  WGS 84
#> # A tibble: 4 × 14
#>     idx osm_type  osm_id osm_key  osm_value     type  postcode countrycode name 
#>   <int> <chr>      <int> <chr>    <chr>         <chr> <chr>    <chr>       <chr>
#> 1     1 R        2706247 place    village       city  71150    FR          Pari…
#> 2     2 R        3464353 place    city          city  <NA>     CN          Shen…
#> 3     3 R         301120 place    municipality  city  59460-0… BR          São …
#> 4     4 R         388103 boundary administrati… city  <NA>     CD          Kins…
#> # ℹ 5 more variables: county <chr>, state <chr>, country <chr>, extent <list>,
#> #   geometry <POINT [°]>

Reverse geocoding means taking point geometries and returning their addresses or place names.

cities3 <- reverse(cities1$geometry, layer = "city")
cities3
#> Simple feature collection with 4 features and 12 fields
#> Geometry type: POINT
#> Dimension:     XY
#> Bounding box:  xmin: -46.63338 ymin: -23.55065 xmax: 114.0545 ymax: 48.8535
#> Geodetic CRS:  WGS 84
#> # A tibble: 4 × 13
#>     idx osm_type  osm_id osm_key osm_value type  countrycode name  state country
#>   <int> <chr>      <int> <chr>   <chr>     <chr> <chr>       <chr> <chr> <chr>  
#> 1     1 R          71525 place   city      city  FR          Paris Isla… France 
#> 2     2 R        3464353 place   city      city  CN          Shen… Guan… China  
#> 3     3 R         298285 place   municipa… city  BR          São … São … Brazil 
#> 4     4 R         388103 bounda… administ… city  CD          Kins… Kins… Democr…
#> # ℹ 3 more variables: extent <list>, county <chr>, geometry <POINT [°]>
all.equal(cities1, cities3)
#> [1] TRUE

Offline geocoding

photon is designed to facilitate offline geocoding. new_photon() can install photon locally. The following code would install and start photon covering the country of Germany in the current working directory.

photon <- new_photon(path = "./photon", region = "Germany")
photon$start()
  • The {photon} package by Timothée Giraud interfaces photon but does not allow the setup of local instances and was abandoned a while ago.
  • The {revgeo} package by Michael Hudecheck implements reverse geocoding using (among others) photon.
  • The {tidygeocoder} and {nominatimlite} packages implement general (OSM) geocoding using web APIs.