rgbif (version 1.0.2)

map_fetch: Fetch aggregated density maps of GBIF occurrences

Description

This function is a wrapper for the GBIF mapping api version 2.0. The mapping API is a web map tile service making it straightforward to visualize GBIF content on interactive maps, and overlay content from other sources. It returns tile maps or vector maps with number of GBIF records per area unit that can be used in a variety of ways, for example in interactive leaflet web maps. Map details are specified by a number of query parameters, some of them optional. Full documentation of the GBIF mapping api can be found at https://www.gbif.org/developer/maps

Usage

map_fetch(source = "density", x = 0, y = 0, z = 0, format = "@1x.png",
  srs = "EPSG:4326", bin = NULL, hexPerTile = NULL, squareSize = NULL,
  style = "classic.point", search = NULL, id = NULL, year = NULL,
  basisOfRecord = NULL, ...)

Arguments

source

(character) Either density for fast, precalculated tiles, or adhoc for any search. Default: density

x

(integer) the zoom. Default: 0

y

(integer) the column. Default: 0

z

(integer) the row. Default: 0

format

(character) The data format, one of:

• .mvt for a vector tile

• @Hx.png for a 256px raster tile

• @1x.png for a 512px raster tile (the default)

• @2x.png for a 1024px raster tile

• @3x.png for a 2048px raster tile

• @4x.png for a 4096px raster tile

srs

(character) Spatial reference system. One of:

• EPSG:3857 (Web Mercator)

• EPSG:4326 (WGS84 plate care?)

• EPSG:3575 (Arctic LAEA on 10 degrees E)

• EPSG:3031 (Antarctic stereographic)

bin

(character) square or hex to aggregate occurrence counts into squares or hexagons. Points by default. optional

hexPerTile

(integer) sets the size of the hexagons (the number horizontally across a tile). optional

squareSize

(integer) sets the size of the squares. Choose a factor of 4096 so they tessalate correctly: probably from 8, 16, 32, 64, 128, 256, 512. optional

style

(character) for raster tiles, choose from the available styles. Defaults to classic.point. optional. THESE DON'T WORK YET.

search

(character) defines what type of subset of all GBIF data to return. Should be one of "taxonKey", "datasetKey", "country", "publisher", "publishingCountry". Without any search parameter, all occurrences will be returned. optional

id

(character) defines the value to be used as filter criterium in the category supplied by search. Appropriate values depend on the search category that is used, for example integer for search = "taxonKey". Has to be provided if search parameter is specified. optional

year

(integer) integer that limits the search to a certain year or, if passing a vector of integers, multiple years, for example 1984 or c(2016, 2017, 2018). optional

basisOfRecord

(character) one or more basis of record states to include records with that basis of record. The full list is: c("OBSERVATION", "HUMAN_OBSERVATION", "MACHINE_OBSERVATION", "MATERIAL_SAMPLE", "PRESERVED_SPECIMEN", "FOSSIL_SPECIMEN", "LIVING_SPECIMEN", "LITERATURE", "UNKNOWN"). optional

...

curl options passed on to crul::HttpClient

Value

an object of class RasterLayer if png format used, or raw bytes when mvt format chosen

Details

This function uses the arguments passed on to generate a query to the GBIF web map API. The API returns a web tile object as png that is read and converted into an R raster object. The break values or nbreaks generate a custom colour palette for the web tile, with each bin corresponding to one grey value. After retrieval, the raster is reclassified to the actual break values. This is a somewhat hacky but nonetheless functional solution in the absence of a GBIF raster API implementation.

We add extent and set the projection for the output. You can reproject after retrieving the output.

References

https://www.gbif.org/developer/maps

Examples

# NOT RUN {
if (
 requireNamespace("png", quietly = TRUE) && 
 requireNamespace("raster", quietly = TRUE)
) {
  x <- map_fetch(search = "taxonKey", id = 3118771, year = 2010)
  x
  # gives a RasterLayer object
  class(x)
  # visualize
  library(raster)
  plot(x)

  # different srs
  ## 3857
  y <- map_fetch(search = "taxonKey", id = 3118771, year = 2010, srs = "EPSG:3857")
  plot(y)
  ## 3031
  z <- map_fetch(search = "taxonKey", id = 3118771, year = 2010, srs = "EPSG:3031")
  plot(z)
  # 3575
  z <- map_fetch(search = "taxonKey", id = 3118771, year = 2010, srs = "EPSG:3575")
  plot(z)

  # bin
  plot(map_fetch(search = "taxonKey", id = 212, year = 1998, bin = "hex",
     hexPerTile = 30, style = "classic-noborder.poly"))

  # map vector tile, gives back raw bytes
  x <- map_fetch(search = "taxonKey", id = 3118771, year = 2010,
    format = ".mvt")
  x[1:10]
 }
# }