Geospatial Search with ArangoSearch

ArangoSearch supports geospatial queries like finding locations and GeoJSON shapes within a radius or area

ArangoSearch can accelerate various types of geospatial queries for data that is indexed by a View. The regular geospatial index can do most of this too, but ArangoSearch allows you to combine geospatial requests with other kinds of searches, like full-text search.

Creating geospatial Analyzers

Geospatial data that can be indexed:

GeoJSON features such as Points and Polygons (with coordinates in [longitude, latitude] order), for example:

{
  "location": {
    "type": "Point",
    "coordinates": [ -73.983, 40.764 ]
  }
}

Coordinates using an array with two numbers in [longitude, latitude] order, for example:
```
{
  "location": [ -73.983, 40.764 ]
}
```
Coordinates using an array with two numbers in [latitude, longitude] order, for example:
```
{
  "location": [ 40.764, -73.983 ]
}
```

Coordinates using two separate numeric attributes, for example:

{
  "location": {
    "lat": 40.764,
    "lng": -73.983
  }
}

You need to create Geo Analyzers manually. There are no pre-configured (built-in) Geo Analyzers.

The data needs to be pre-processed with a geojson or geo_s2 Analyzer in case of GeoJSON or coordinate arrays in [longitude, latitude] order.
For coordinate arrays in [latitude, longitude] order or coordinate pairs using separate attributes, you need to use a geopoint Analyzer.

Custom Analyzers:

Create a geojson Analyzer in arangosh to pre-process arbitrary GeoJSON features or [longitude, latitude] arrays. The default properties are usually what you want, therefore an empty object is passed:

//db._useDatabase("your_database"); // Analyzer will be created in current database
var analyzers = require("@arangodb/analyzers");
analyzers.save("geojson", "geojson", {}, ["frequency", "norm", "position"]);

See geojson Analyzer for details.

In the Enterprise Edition, you can use the geo_s2 Analyzer instead of the geojson Analyzer to more efficiently index geo-spatial data. It is mostly a drop-in replacement, but you can choose between different binary formats. See Analyzers for details.

Create a geopoint Analyzer in arangosh using the default properties (empty object) to pre-process coordinate arrays in [latitude, longitude] order:

//db._useDatabase("your_database"); // Analyzer will be created in current database
var analyzers = require("@arangodb/analyzers");
analyzers.save("geo_pair", "geopoint", {}, ["frequency", "norm", "position"]);

Create a geopoint Analyzer in arangosh to pre-process coordinates with latitude and longitude stored in two different attributes. These attributes cannot be at the top-level of the document, but must be nested in an object, e.g. { location: { lat: 40.78, lon: -73.97 } }. The path relative to the parent attribute (here: location) needs to be described in the Analyzer properties for each of the coordinate attributes:

//db._useDatabase("your_database"); // Analyzer will be created in current database
var analyzers = require("@arangodb/analyzers");
analyzers.save("geo_latlng", "geopoint", { latitude: ["lat"], longitude: ["lng"] }, ["frequency", "norm", "position"]);

Using the example dataset

Load the dataset into an ArangoDB instance and create a View restaurantsViews as described below:

Dataset: Demo Geo S2 dataset

View definition

`search-alias` View

db.restaurants.ensureIndex({ name: "inv-rest", type: "inverted", fields: [ { name: "location", analyzer: "geojson" } ] });
db.neighborhoods.ensureIndex({ name: "inv-hood", type: "inverted", fields: [ "name", { name: "geometry", analyzer: "geojson" } ] });
db._createView("restaurantsViewAlias", "search-alias", { indexes: [
  { collection: "restaurants", index: "inv-rest" },
  { collection: "neighborhoods", index: "inv-hood" }
] });

`arangosearch` View

{
  "links": {
    "restaurants": {
      "fields": {
        "location": {
          "analyzers": [
            "geojson"
          ]
        }
      }
    }
  },
  "neighborhoods": {
    "fields": {
      "name": {
        "analyzers": [
          "identity"
        ]
      },
      "geometry": {
        "analyzers": [
          "geojson"
        ]
      }
    }
  }
}

Search for points within a radius

Using the Museum of Modern Arts as reference location, find restaurants within a 100 meter radius. Return the matches sorted by distance and include how far away they are from the reference point in the result.

search-alias View:

LET moma = GEO_POINT(-73.983, 40.764)
FOR doc IN restaurantsViewAlias
  SEARCH GEO_DISTANCE(doc.location, moma) < 100
  LET distance = GEO_DISTANCE(doc.location, moma)
  SORT distance
  RETURN {
    geometry: doc.location,
    distance
  }

arangosearch View:

LET moma = GEO_POINT(-73.983, 40.764)
FOR doc IN restaurantsView
  SEARCH ANALYZER(GEO_DISTANCE(doc.location, moma) < 100, "geojson")
  LET distance = GEO_DISTANCE(doc.location, moma)
  SORT distance
  RETURN {
    geometry: doc.location,
    distance
  }

Search for restaurants with Cafe in their name within a radius of 1000 meters and return the ten closest matches:

search-alias View:

LET moma = GEO_POINT(-73.983, 40.764)
FOR doc IN restaurantsViewAlias
  SEARCH LIKE(doc.name, "%Cafe%")
     AND GEO_DISTANCE(doc.location, moma) < 1000
  LET distance = GEO_DISTANCE(doc.location, moma)
  SORT distance
  LIMIT 10
  RETURN {
    geometry: doc.location,
    name: doc.name,
    distance
  }

arangosearch View:

LET moma = GEO_POINT(-73.983, 40.764)
FOR doc IN restaurantsView
  SEARCH LIKE(doc.name, "%Cafe%")
     AND ANALYZER(GEO_DISTANCE(doc.location, moma) < 1000, "geojson")
  LET distance = GEO_DISTANCE(doc.location, moma)
  SORT distance
  LIMIT 10
  RETURN {
    geometry: doc.location,
    name: doc.name,
    distance
  }

Search for points within a polygon

First off, search for the neighborhood Upper West Side in a subquery and return its GeoJSON Polygon. Then search for restaurants that are contained in this polygon and return them together with the polygon itself:

search-alias View:

LET upperWestSide = FIRST(
  FOR doc IN restaurantsViewAlias
    SEARCH doc.name == "Upper West Side"
    RETURN doc.geometry
)
FOR result IN PUSH(
  FOR doc IN restaurantsViewAlias
    SEARCH GEO_CONTAINS(upperWestSide, doc.location)
    RETURN doc.location,
  upperWestSide
)
  RETURN result

arangosearch View:

LET upperWestSide = FIRST(
  FOR doc IN restaurantsView
    SEARCH doc.name == "Upper West Side"
    RETURN doc.geometry
)
FOR result IN PUSH(
  FOR doc IN restaurantsView
    SEARCH ANALYZER(GEO_CONTAINS(upperWestSide, doc.location), "geojson")
    RETURN doc.location,
  upperWestSide
)
  RETURN result

ArangoSearch geospatial query for points in a polygon

You do not have to look up the polygon, you can also provide one inline. It is also not necessary to return the polygon, you can return the matches only:

LET upperWestSide = {
  "coordinates": [
    [
      [-73.9600301843709, 40.79803810789689], [-73.96052271669541, 40.797368469462334],
      [-73.96097971807933, 40.79673864404529], [-73.96144060655736, 40.79611082718394],
      [-73.96189985460951, 40.79547927006112], [-73.96235980150668, 40.79485206056065],
      [-73.96280590635729, 40.79423581323211], [-73.96371096541819, 40.79301293488322],
      [-73.9641759852132, 40.79236204502772], [-73.96468540739478, 40.79166402679883],
      [-73.96517705499011, 40.79099034109932], [-73.96562799538655, 40.790366117129004],
      [-73.96609500572444, 40.78973438976665], [-73.96655226678917, 40.78910715282553],
      [-73.96700977073398, 40.78847679074218], [-73.96744908373155, 40.78786072059045],
      [-73.96792696354466, 40.78722157112602], [-73.96838479313664, 40.78659569652393],
      [-73.96884378957469, 40.78596738856434], [-73.96933573318945, 40.78529327955705],
      [-73.96983225556819, 40.7846109105862], [-73.97030068162124, 40.78397541394847],
      [-73.97076013116715, 40.783340137553594], [-73.97122292220932, 40.782706256089995],
      [-73.9717230555586, 40.78202147595964], [-73.97357117423289, 40.7794778616211],
      [-73.97406668257638, 40.77880541672153], [-73.97453231422314, 40.77816778452296],
      [-73.97499744020544, 40.777532546222], [-73.97783054404911, 40.77872973181589],
      [-73.98067365344895, 40.7799251824873], [-73.98140948736065, 40.780235418619405],
      [-73.98151911347311, 40.78028175751621], [-73.9816278736105, 40.780328934969766],
      [-73.98232616371553, 40.78062377270337], [-73.9835260146705, 40.781130011022704],
      [-73.98507184345014, 40.781779680969194], [-73.98536952677372, 40.781078372362586],
      [-73.98567936117642, 40.78031263333493], [-73.98654378951805, 40.780657980791055],
      [-73.98707137465644, 40.78090638159226], [-73.98730772854313, 40.781041303287786],
      [-73.98736363983177, 40.78106280511045], [-73.98741432690473, 40.7810875110951],
      [-73.98746219857024, 40.7811086095956], [-73.98799363156404, 40.78134281734761],
      [-73.98812746102577, 40.78140179644223], [-73.98804128806725, 40.78158596085119],
      [-73.9881002938246, 40.78160287830527], [-73.98807644914505, 40.78165093500162],
      [-73.98801805997222, 40.78163418881042], [-73.98796079284213, 40.781770987031514],
      [-73.98791968459247, 40.78183347771321], [-73.98787728725019, 40.78189205083216],
      [-73.98647480368592, 40.783916573718706], [-73.98625187003543, 40.78423876424543],
      [-73.98611372725294, 40.78443891187735], [-73.98561580396368, 40.78514186259503],
      [-73.98546581197026, 40.78536070057543], [-73.98617270496544, 40.786068452258675],
      [-73.98645586240198, 40.7859192190814], [-73.98707234561569, 40.78518963831753],
      [-73.98711901394266, 40.78521031850151], [-73.98649778102359, 40.78595120288725],
      [-73.98616462880626, 40.786121882448306], [-73.98612842248588, 40.78623900133112],
      [-73.98607113521973, 40.78624070602659], [-73.98602727478911, 40.78622896423671],
      [-73.98609763784941, 40.786058225697936], [-73.98542932126942, 40.78541394218462],
      [-73.98508113773205, 40.785921935110444], [-73.98519883325449, 40.785966552197756],
      [-73.98517050238989, 40.786013334158156], [-73.98521621867376, 40.786030501313824],
      [-73.98525509797992, 40.78597620551157], [-73.98524273937655, 40.78597257215073],
      [-73.98524962933016, 40.78596313985583], [-73.98528177918672, 40.785978620950054],
      [-73.9852400328845, 40.786035858136785], [-73.98568388524215, 40.78622212391968],
      [-73.98571752900456, 40.78617599466878], [-73.98576566029752, 40.786196274858625],
      [-73.9856828719225, 40.78630978621313], [-73.98563627093053, 40.786290150146684],
      [-73.98567072256468, 40.786242911993796], [-73.98561523764435, 40.78621964571528],
      [-73.98520511880037, 40.78604766921276], [-73.98521103560748, 40.78603955488367],
      [-73.98516263994709, 40.78602099926717], [-73.98513163631205, 40.786060297019965],
      [-73.98501696406497, 40.78601423719563], [-73.98493597820354, 40.786130720650974],
      [-73.98465507883022, 40.78653474180794], [-73.98574378790113, 40.78657008235215],
      [-73.98589227228327, 40.78642652901958], [-73.98594285499497, 40.78645284788032],
      [-73.98594956155667, 40.786487113463934], [-73.98581237352651, 40.78661686535709],
      [-73.98513520970327, 40.7865876183929], [-73.98461942858408, 40.78658601634982],
      [-73.98369521623664, 40.7879152813127], [-73.98234664147564, 40.789854780772636],
      [-73.98188645946546, 40.79051658043251], [-73.98139174468567, 40.79122824550256],
      [-73.9812893737095, 40.79137550943302], [-73.9809470835408, 40.79186789327993],
      [-73.980537679464, 40.79245681262498], [-73.98043434256003, 40.79259428309673],
      [-73.98013222578662, 40.79299376538315], [-73.98004684398002, 40.79311352516391],
      [-73.9792882208298, 40.79417763216331], [-73.97828949152755, 40.79558676046088],
      [-73.97779475503205, 40.79628462977189], [-73.97685207845194, 40.79763134839318],
      [-73.97639951930574, 40.79827321018994], [-73.97628527884252, 40.798435235410054],
      [-73.97583055785255, 40.799092280410655], [-73.97578169321191, 40.799156256780286],
      [-73.97579140130195, 40.799209627886206], [-73.97576219486481, 40.79926017354928],
      [-73.97554385822018, 40.79952825063732], [-73.97526783234453, 40.79993284953172],
      [-73.97508668067891, 40.800189533632995], [-73.97496436808184, 40.80036963419388],
      [-73.97483924436003, 40.800558243262664], [-73.97466556722725, 40.80081351473415],
      [-73.97448722520987, 40.801057428896804], [-73.97414361823468, 40.80151689534114],
      [-73.97394098366709, 40.801809025361415], [-73.97389989052462, 40.80188986353119],
      [-73.97377477246009, 40.802045845948555], [-73.97372060455763, 40.80216781528022],
      [-73.97361322463904, 40.80229685988716], [-73.9735422772157, 40.802356411250294],
      [-73.97336671067801, 40.80263011334645], [-73.97320518045738, 40.802830058276285],
      [-73.97312859120993, 40.80297471550862], [-73.97307070537943, 40.8030555484474],
      [-73.97303522902072, 40.803073973741895], [-73.97292317001968, 40.80324982284384],
      [-73.97286807262155, 40.80332073417601], [-73.97287179081519, 40.80335618764528],
      [-73.9727990775659, 40.803329159656634], [-73.9726574474597, 40.803276514162725],
      [-73.97257779806121, 40.803247183771205], [-73.97250022180596, 40.80321661262814],
      [-73.97150381003809, 40.80283773617443], [-73.97032589767365, 40.802384560870536],
      [-73.9702740046587, 40.80235903699402], [-73.97021594793262, 40.80233585148787],
      [-73.9700474216526, 40.8022650103398], [-73.9685836074654, 40.80163547026629],
      [-73.96798415954912, 40.80139826627661], [-73.967873797219, 40.801351698184384],
      [-73.96775900356977, 40.80130351598543], [-73.96571144280439, 40.80043806998765],
      [-73.96286980146162, 40.79923967661966], [-73.96147779901374, 40.79865415643638],
      [-73.9600301843709, 40.79803810789689]
    ]
  ],
  "type": "Polygon"
}
FOR doc IN restaurantsView
  SEARCH ANALYZER(GEO_CONTAINS(upperWestSide, doc.location), "geojson")
  RETURN doc.location

/* `search-alias` View:
FOR doc IN restaurantsViewAlias
  SEARCH GEO_CONTAINS(upperWestSide, doc.location)
  RETURN doc.location
*/

Search for polygons within polygons

Define a GeoJSON polygon that is a rectangle, then search for neighborhoods that are fully contained in this area:

LET sides = {
  left: -74,
  top: 40.8,
  right: -73.93,
  bottom: 40.76
}

LET rect = GEO_POLYGON([
  [sides.left, sides.bottom],
  [sides.left, sides.top],
  [sides.right, sides.top],
  [sides.right, sides.bottom],
  [sides.left, sides.bottom]
])

FOR result IN PUSH(
  FOR doc IN restaurantsView
    SEARCH ANALYZER(GEO_CONTAINS(rect, doc.geometry), "geojson")
  /* `search-alias` View:
  FOR doc IN restaurantsViewAlias
    SEARCH GEO_CONTAINS(rect, doc.geometry)
  */
    RETURN doc.geometry,
  rect
)
  RETURN result

ArangoSearch geosptial query for polygons in a polygon

Searching for geo features in a rectangle is something you can use together with an interactive map that the user can select the area of interest with. Take a look at the lunch break video about the ArangoBnB demo project to learn more.

Search for polygons intersecting polygons

Define a GeoJSON polygon that is a rectangle, then search for neighborhoods that intersect with this area:

LET sides = {
  left: -74,
  top: 40.8,
  right: -73.93,
  bottom: 40.76
}

LET rect = GEO_POLYGON([
  [sides.left, sides.bottom],
  [sides.left, sides.top],
  [sides.right, sides.top],
  [sides.right, sides.bottom],
  [sides.left, sides.bottom]
])

FOR result IN PUSH(
  FOR doc IN restaurantsView
    SEARCH ANALYZER(GEO_INTERSECTS(rect, doc.geometry), "geojson")
  /* `search-alias` View:
  FOR doc IN restaurantsViewAlias
    SEARCH GEO_INTERSECTS(rect, doc.geometry)
  */
    RETURN doc.geometry,
  rect
)
  RETURN result

ArangoSearch geospatial query for polygons intersecting a polygon

❮ Fuzzy search Search highlighting ❯

Geospatial Search with ArangoSearch

Creating geospatial Analyzers

Using the example dataset

View definition

search-alias View

arangosearch View

Search for points within a radius

Search for points within a polygon

Search for polygons within polygons

Search for polygons intersecting polygons

`search-alias` View

`arangosearch` View