geojsonsf
geojsonsf
A simple, low-dependency and fast converter between GeoJSON and Simple Feature objects in R.
v1.0
Converts
- GeoJSON –>
sf - GeoJSON –>
sfc sf–> GeoJSONsfc–> GeoJSON- GeoJSON –> Well-known text
As per GeoJSON (RFC 7946
specification), foreign
members are ignored, and nested objects and arrays inside the
properties object are converted to string/characters.
Installation
When released on CRAN you install it in the usual way
install.packages("geojsonsf")Install the development version from GitHub with
# install.packages("devtools")
devtools::install_github("SymbolixAU/geojsonsf")Motivation
To quickly parse between GeoJSON and sf objects, and to handle cases
not supported by sf, e.g. arrays of geometries
Arrays of GeoJSON
js <- '[
{
"type": "FeatureCollection",
"features": [
{
"type": "Feature",
"properties": {"id":1,"val":true},
"geometry": {"type": "Point", "coordinates": [100.0, 0.0]}
},
{
"type": "Feature",
"properties": {"id":2,"val":false},
"geometry": {"type": "LineString", "coordinates": [[201.0, 0.0], [102.0, 1.0]]}
},
{
"type": "Feature",
"properties": {"id":3},
"geometry": {"type": "LineString", "coordinates": [[301.0, 0.0], [102.0, 1.0]]}
}
]
},
{
"type": "FeatureCollection",
"features": [
{
"type": "Feature",
"properties": {"id":1},
"geometry": {"type": "Point", "coordinates": [100.0, 0.0]}
},
{
"type": "Feature",
"properties": {"val":false},
"geometry": {"type": "LineString", "coordinates": [[501.0, 0.0], [102.0, 1.0]]}
},
{
"type": "Feature",
"properties": {"hello":"world"},
"geometry": {"type": "LineString", "coordinates": [[601.0, 0.0], [102.0, 1.0]]}
}
]
}
]'
sf <- geojson_sf(js)
sf
# Simple feature collection with 6 features and 3 fields
# geometry type: GEOMETRY
# dimension: XY
# bbox: xmin: 100 ymin: 0 xmax: 601 ymax: 1
# epsg (SRID): 4326
# proj4string: +proj=longlat +datum=WGS84 +no_defs
# geometry hello id val
# 1 POINT (100 0) <NA> 1 1
# 2 LINESTRING (201 0, 102 1) <NA> 2 0
# 3 LINESTRING (301 0, 102 1) <NA> 3 <NA>
# 4 POINT (100 0) <NA> 1 <NA>
# 5 LINESTRING (501 0, 102 1) <NA> NA 0
# 6 LINESTRING (601 0, 102 1) world NA <NA>And back again to GeoJSON
js <- sf_geojson(sf)
jsonlite::prettify(js)
# {
# "type": "FeatureCollection",
# "features": [
# {
# "type": "Feature",
# "properties": {
# "hello": null,
# "id": 1,
# "val": "1"
# },
# "geometry": {
# "type": "Point",
# "coordinates": [
# 100,
# 0
# ]
# }
# },
# {
# "type": "Feature",
# "properties": {
# "hello": null,
# "id": 2,
# "val": "0"
# },
# "geometry": {
# "type": "LineString",
# "coordinates": [
# [
# 201,
# 0
# ],
# [
# 102,
# 1
# ]
# ]
# }
# },
# {
# "type": "Feature",
# "properties": {
# "hello": null,
# "id": 3,
# "val": null
# },
# "geometry": {
# "type": "LineString",
# "coordinates": [
# [
# 301,
# 0
# ],
# [
# 102,
# 1
# ]
# ]
# }
# },
# {
# "type": "Feature",
# "properties": {
# "hello": null,
# "id": 1,
# "val": null
# },
# "geometry": {
# "type": "Point",
# "coordinates": [
# 100,
# 0
# ]
# }
# },
# {
# "type": "Feature",
# "properties": {
# "hello": null,
# "id": null,
# "val": "0"
# },
# "geometry": {
# "type": "LineString",
# "coordinates": [
# [
# 501,
# 0
# ],
# [
# 102,
# 1
# ]
# ]
# }
# },
# {
# "type": "Feature",
# "properties": {
# "hello": "world",
# "id": null,
# "val": null
# },
# "geometry": {
# "type": "LineString",
# "coordinates": [
# [
# 601,
# 0
# ],
# [
# 102,
# 1
# ]
# ]
# }
# }
# ]
# }
# Speed
This benchmark shows a comparison with library(sf) for converting a
string of GeoJSON of 3,221 counties in the US in to an sf
object
myurl <- "http://eric.clst.org/assets/wiki/uploads/Stuff/gz_2010_us_050_00_500k.json"
geo <- readLines(myurl)
geo <- paste0(geo, collapse = "")
library(microbenchmark)
microbenchmark(
geojsonsf = {
geojson_sf(geo)
},
sf = {
sf::st_read(geo, quiet = T)
},
times = 2
)
# Unit: milliseconds
# expr min lq mean median uq max neval
# geojsonsf 681.376 681.376 718.0487 718.0487 754.7214 754.7214 2
# sf 1814.307 1814.307 1816.5013 1816.5013 1818.6955 1818.6955 2Reading directly from a URL is comparable between the two
myurl <- "http://eric.clst.org/assets/wiki/uploads/Stuff/gz_2010_us_050_00_500k.json"
library(microbenchmark)
microbenchmark(
geojsonsf = {
geojson_sf(myurl)
},
sf = {
sf::st_read(myurl, quiet = T)
},
times = 2
)
# Unit: seconds
# expr min lq mean median uq max neval
# geojsonsf 6.617584 6.617584 6.680125 6.680125 6.742666 6.742666 2
# sf 6.914326 6.914326 7.672934 7.672934 8.431542 8.431542 2library(rgdal)
microbenchmark(
gdal = {
readOGR(myurl, "OGRGeoJSON")
},
geojsonsf = {
myurl <- "http://eric.clst.org/assets/wiki/uploads/Stuff/gz_2010_us_050_00_500k.json"
geo <- readLines(myurl)
geo <- paste0(geo, collapse = "")
geojson_sf(geo)
},
times = 5
)
# expr min lq mean median uq max neval
# gdal 58.51037 60.05683 66.33925 65.07506 72.08371 75.97028 5
# geojsonsf 11.91515 13.37422 14.02232 13.88782 14.61826 16.31612 5A visual check to see both objects are the same
library(googleway)
gsf <- geojson_sf(geo)
google_map() %>%
add_polygons(gsf[!gsf$STATE %in% c("02","15","72"), ],
fill_colour = "CENSUSAREA",
stroke_weight = 0)sf <- sf::st_read(geo, quiet = T)
plot(st_geometry(sf[!sf$STATE %in% c("02", "15", "72"), ]))