|Modifier and Type||Method and Description|
public static TopFieldDocs nearest(IndexSearcher searcher, String field, double latitude, double longitude, int n) throws IOException
nnearest indexed points to the provided point, according to Haversine distance.
This is functionally equivalent to running
MatchAllDocsQuery with a
LatLonDocValuesField.newDistanceSort(java.lang.String, double, double),
but is far more efficient since it takes advantage of properties the indexed BKD tree. Currently this
only works with
Lucene86PointsFormat (used by the default codec). Multi-valued fields are
currently not de-duplicated, so if a document had multiple instances of the specified field that
make it into the top n, that document will appear more than once.
Documents are ordered by ascending distance from the location. The value returned in
the hits contains a Double instance with the distance in meters.
searcher- IndexSearcher to find nearest points from.
field- field name. must not be null.
latitude- latitude at the center: must be within standard +/-90 coordinate bounds.
longitude- longitude at the center: must be within standard +/-180 coordinate bounds.
n- the number of nearest neighbors to retrieve.
FieldDocis the distance in meters
IllegalArgumentException- if the underlying PointValues is not a
Lucene60PointsReader(this is a current limitation), or if
searcheris null, or if
IOException- if an IOException occurs while finding the points.
Copyright © 2000-2021 Apache Software Foundation. All Rights Reserved.