geoarrow

The goal of geoarrow is to leverage the features of the arrow package and larger Apache Arrow ecosystem for geospatial data. The geoarrow package provides an R implementation of the GeoParquet file format of and the draft geoarrow data specification, defining extension array types for vector geospatial data.

Installation

You can install the released version of geoarrow from CRAN with:

install.packages("geoarrow")

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

# install.packages("pak")
pak::pak("geoarrow/geoarrow-r")

Example

The geoarrow package implements conversions to/from various geospatial types (e.g., sf, sfc, s2, wk) with various Arrow representations (e.g., arrow, nanoarrow). The most useful conversions are between the arrow and sf packages, which in most cases allow sf objects to be passed to arrow functions directly after library(geoarrow) or requireNamespace("geoarrow") has been called.

library(geoarrow)
library(arrow, warn.conflicts = FALSE)
#> Warning: package 'arrow' was built under R version 4.3.3
library(sf)
#> Linking to GEOS 3.11.0, GDAL 3.5.3, PROJ 9.1.0; sf_use_s2() is TRUE

nc <- read_sf(system.file("gpkg/nc.gpkg", package = "sf"))
tf <- tempfile(fileext = ".parquet")

nc |> 
  tibble::as_tibble() |> 
  write_parquet(tf)

open_dataset(tf) |> 
  dplyr::filter(startsWith(NAME, "A")) |>
  dplyr::select(NAME, geom) |> 
  st_as_sf()
#> Simple feature collection with 6 features and 1 field
#> Geometry type: MULTIPOLYGON
#> Dimension:     XY
#> Bounding box:  xmin: -82.07776 ymin: 34.80792 xmax: -79.23799 ymax: 36.58965
#> Geodetic CRS:  NAD27
#>        NAME                           geom
#> 1      Ashe MULTIPOLYGON (((-81.47276 3...
#> 2 Alleghany MULTIPOLYGON (((-81.23989 3...
#> 3     Avery MULTIPOLYGON (((-81.94135 3...
#> 4  Alamance MULTIPOLYGON (((-79.24619 3...
#> 5 Alexander MULTIPOLYGON (((-81.10889 3...
#> 6     Anson MULTIPOLYGON (((-79.91995 3...

By default, arrow objects are converted to a neutral wrapper around chunked Arrow memory, which in turn implements conversions to most spatial types:

df <- read_parquet(tf)
df$geom
#> <geoarrow_vctr geoarrow.multipolygon{list}[100]>
#>  [1] <MULTIPOLYGON (((-81.4727554 36.2343559, -81.5408401 36.2725067, -81.56>
#>  [2] <MULTIPOLYGON (((-81.2398911 36.3653641, -81.2406921 36.3794174, -81.26>
#>  [3] <MULTIPOLYGON (((-80.4563446 36.2425575, -80.476387 36.2547264, -80.536>
#>  [4] <MULTIPOLYGON (((-76.0089722 36.3195953, -76.0173492 36.3377304, -76.03>
#>  [5] <MULTIPOLYGON (((-77.2176666 36.2409821, -77.2346115 36.2145996, -77.29>
#>  [6] <MULTIPOLYGON (((-76.7450638 36.2339172, -76.98069 36.2302361, -76.9947>
#>  [7] <MULTIPOLYGON (((-76.0089722 36.3195953, -75.9571838 36.1937714, -75.98>
#>  [8] <MULTIPOLYGON (((-76.5625076 36.3405685, -76.6042404 36.3149834, -76.64>
#>  [9] <MULTIPOLYGON (((-78.3087616 36.2600403, -78.2829285 36.2918816, -78.32>
#> [10] <MULTIPOLYGON (((-80.0256729 36.2502327, -80.4530106 36.2570877, -80.43>
#> [11] <MULTIPOLYGON (((-79.5305099 36.2461357, -79.5102997 36.547657, -79.217>
#> [12] <MULTIPOLYGON (((-79.5305099 36.2461357, -79.5305786 36.2361565, -80.02>
#> [13] <MULTIPOLYGON (((-78.7491226 36.063591, -78.788414 36.0621834, -78.8040>
#> [14] <MULTIPOLYGON (((-78.8068008 36.231575, -78.9510803 36.2338371, -79.159>
#> [15] <MULTIPOLYGON (((-78.4925232 36.1735878, -78.5147171 36.1752243, -78.51>
#> [16] <MULTIPOLYGON (((-77.3322067 36.0679817, -77.4053116 35.9947166, -77.42>
#> [17] <MULTIPOLYGON (((-76.2989273 36.2142296, -76.324234 36.2336235, -76.372>
#> [18] <MULTIPOLYGON (((-81.0205688 36.034935, -81.0840836 36.0207672, -81.124>
#> [19] <MULTIPOLYGON (((-81.806221 36.1045609, -81.8171539 36.1093864, -81.822>
#> [20] <MULTIPOLYGON (((-76.4805298 36.079792, -76.5369568 36.087925, -76.5755>
#> ...and 80 more values
st_as_sfc(df$geom)
#> Geometry set for 100 features 
#> Geometry type: MULTIPOLYGON
#> Dimension:     XY
#> Bounding box:  xmin: -84.32385 ymin: 33.88199 xmax: -75.45698 ymax: 36.58965
#> Geodetic CRS:  NAD27
#> First 5 geometries:
#> MULTIPOLYGON (((-81.47276 36.23436, -81.54084 3...
#> MULTIPOLYGON (((-81.23989 36.36536, -81.24069 3...
#> MULTIPOLYGON (((-80.45634 36.24256, -80.47639 3...
#> MULTIPOLYGON (((-76.00897 36.3196, -76.01735 36...
#> MULTIPOLYGON (((-77.21767 36.24098, -77.23461 3...

The entry point to creating arrays is as_geoarrow_vctr():

as_geoarrow_vctr(c("POINT (0 1)", "POINT (2 3)"))
#> <geoarrow_vctr geoarrow.wkt{string}[2]>
#> [1] <POINT (0 1)> <POINT (2 3)>

By default these do not attempt to create a new storage type; however, you can request a storage type or infer one from the data:

as_geoarrow_vctr(c("POINT (0 1)", "POINT (2 3)"), schema = geoarrow_native("POINT"))
#> <geoarrow_vctr geoarrow.point{struct}[2]>
#> [1] <POINT (0 1)> <POINT (2 3)>

vctr <- as_geoarrow_vctr(c("POINT (0 1)", "POINT (2 3)"))
as_geoarrow_vctr(vctr, schema = infer_geoarrow_schema(vctr))
#> <geoarrow_vctr geoarrow.point{struct}[2]>
#> [1] <POINT (0 1)> <POINT (2 3)>

There are a number of files to use as examples at https://geoarrow.org/data that can be read with arrow::read_ipc_file():

url <- "https://github.com/geoarrow/geoarrow-data/releases/download/v0.1.0/ns-water-basin_point.arrow"
tab <- read_ipc_file(url, as_data_frame = FALSE)
tab$geometry$type
#> GeometryExtensionType
#> geoarrow.multipoint <CRS: {"$schema":"https://proj.or...