Use geo queries
Use geospatial queries in MongoDB compatible operations to query documents that exist within a certain range from a specific longitude and latitude.
Before you begin
-
Ensure that you have access to an existing MongoDB compatible operations database, or Create a database and connect to it .
-
Ensure that you have at least one 2dsphere index, or Create a 2dsphere index .
GeoJSON objects
To run a geospatial query against fields in your collection, the fields you query must be GeoJSON objects or GeoPoints .
Run a geospatial query
You can perform a geospatial query using the $near
operator, which calculates
the distance of a geo point relative to geo points within
documents. These documents are then sorted from nearest to farthest distance in
the query results. You can also override this sort order by defining a secondary sort, other
than $natural
, in your query. The $near
operator must exist in a document
field in the query filter, and must contain a $geometry
GeoJSON field.
In the following example, the $near
operator is used to calculate the distance
between the geo point (-122.084, 37.4221)
and the geo points located in the location
field of all documents in myCollection.
The documents are returned
and ordered by nearest to farthest distance between the two points.
db.myCollection.find ({
location:
{
$near
:
{
$geometry
:
{
type:
'Point'
,
coordinates:
[
-122.084,
37
.4221
]
}
}
}
}
)
You can also use the optional $maxDistance
and $minDistance
fields
to control the distance in meters from the point of your query.
The following example shows a query where returned documents must be at least
500 meters and at most 2000 meters from the point (-122.084, 37.4221)
:
db.myCollection.find ({
location:
{
$near
:
{
$geometry
:
{
type:
'Point'
,
coordinates:
[
-122.084,
37
.4221
]
}
,
$maxDistance
:
2000
,
$minDistance
:
500
}
}
}
If your index is partitioned, then you can filter based on the partition by
including the partition in an "and" equality filter within your query.
For example, if you had a region
partition and wanted to filter your query
results by the midwest
region, you could do the following:
db.myCollection.find (
{
$and
:
[
{
location:
{
$near
:
{
$geometry
:
{
type:
'Point'
,
coordinates:
[
-122.084,
37
.4221
]
}
,
}
}
,
{
"region"
:
"midwest"
}
]
}
)
The value of your partition must be a string. Your partition filter must be
joined to your search query by an $and
operator.
Limitations
-
$nearoperators and$textoperators can't be used in the same the query. -
$nearcan't be nested in a multi-clause$orstatement unless$nearis the only expression in the$orclause. -
$nearcan't be used with the$notor$noroperators in a query. -
$nearisn't supported in aggregation queries.
