osmdata (version 0.2.3)

opq: Build an Overpass query


Build an Overpass query


  bbox = NULL,
  nodes_only = FALSE,
  osm_types = c("node", "way", "relation"),
  out = c("body", "tags", "meta", "skel", "tags center", "ids"),
  datetime = NULL,
  datetime2 = NULL,
  adiff = FALSE,
  timeout = 25,


An overpass_query object



Either (i) four numeric values specifying the maximal and minimal longitudes and latitudes, in the form c(xmin, ymin, xmax, ymax) or (ii) a character string in the form xmin,ymin,xmax,ymax. These will be passed to getbb to be converted to a numerical bounding box. Can also be (iii) a matrix representing a bounding polygon as returned from getbb(..., format_out = "polygon"). To search in an area, (iv) a character string with a relation or a (closed) way id in the format "way(id:1)", "relation(id:1, 2)" or "relation(id:1, 2, 3); way(id:2)" as returned by getbb(..., format_out = "osm_type_id") or bbox_to_string with a data.frame from getbb(..., format_out = "data.frame") to select all areas combined (relations and ways).


WARNING: this parameter is equivalent to osm_types = "node" and will be DEPRECATED. If TRUE, query OSM nodes only. Some OSM structures such as place = "city" or highway = "traffic_signals" are represented by nodes only. Queries are built by default to return all nodes, ways, and relation, but this can be very inefficient for node-only queries. Setting this value to TRUE for such cases makes queries more efficient, with data returned in the osm_points list item.


A character vector with several OSM types to query: node, way and relation is the default. nwr, nw, wr, nr and rel are also valid types. Ignored if nodes_only = TRUE. osm_types = "node" is equivalent to nodes_only = TRUE.


The level of verbosity of the overpass result: body (geometries and tags, the default), tags (tags without geometry), meta (like body + Timestamp, Version, Changeset, User, User ID of the last edition), skel (geometries only), tags center (tags without geometry + the coordinates of the center of the bounding box) and ids (type and id of the objects only).


If specified, a date and time to extract data from the OSM database as it was up to the specified date and time, as described at https://wiki.openstreetmap.org/wiki/Overpass_API/Overpass_QL#date. This must be in ISO8601 format ("YYYY-MM-DDThh:mm:ssZ"), where both the "T" and "Z" characters must be present.


If specified, return the difference in the OSM database between datetime and datetime2, where datetime2 > datetime. See https://wiki.openstreetmap.org/wiki/Overpass_API/Overpass_QL#Difference_between_two_dates_(diff).


If TRUE, query for augmented difference. The result indicates what happened to the modified and deleted OSM objects. Requires datetime(2)*.


It may be necessary to increase this value for large queries, because the server may time out before all data are delivered.


The default memory size for the 'overpass' server in bytes; may need to be increased in order to handle large queries.


The out statement for tags, tags centerand id, do not return geometries. Neither out = "meta" nor adiff = TRUE options are implemented for all osmdata_* functions yet. Use osmdata_xml or osmdata_data_frame to get the result of these queries. See the documentation of the out statement and augmented difference for more details about these options.

See Also

Other queries: add_osm_features(), add_osm_feature(), bbox_to_string(), getbb(), opq_around(), opq_csv(), opq_enclosing(), opq_osm_id(), opq_string(), overpass_status()


Run this code
if (FALSE) {
q <- getbb ("portsmouth", display_name_contains = "United States") %>%
    opq () %>%
    add_osm_feature ("amenity", "restaurant") %>%
    add_osm_feature ("amenity", "pub")
osmdata_sf (q) # all objects that are restaurants AND pubs (there are none!)
q1 <- getbb ("portsmouth", display_name_contains = "United States") %>%
    opq () %>%
    add_osm_feature ("amenity", "restaurant")
q2 <- getbb ("portsmouth", display_name_contains = "United States") %>%
    opq () %>%
    add_osm_feature ("amenity", "pub")
c (osmdata_sf (q1), osmdata_sf (q2)) # all restaurants OR pubs

# Use nodes_only to retrieve single point data only, such as for central
# locations of cities.
opq <- opq (bbox, nodes_only = TRUE) %>%
    add_osm_feature (key = "place", value = "city") %>%
    osmdata_sf (quiet = FALSE)

# Filter by a search area
qa1 <- getbb ("Catalan Countries", format_out = "osm_type_id") %>%
    opq (nodes_only = TRUE) %>%
    add_osm_feature (key = "capital", value = "4")
opqa1 <- osmdata_sf (qa1)
# Filter by a multiple search areas
bb <- getbb ("Vilafranca", format_out = "data.frame")
qa2 <- bbox_to_string (bb [bb$osm_type != "node", ]) %>%
    opq (nodes_only = TRUE) %>%
    add_osm_feature (key = "place")
opqa2 <- osmdata_sf (qa2)

Run the code above in your browser using DataLab