[
https://issues.apache.org/jira/browse/LUCENE-8736?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16800456#comment-16800456
]
Ignacio Vera edited comment on LUCENE-8736 at 3/25/19 8:25 AM:
---------------------------------------------------------------
Thanks [~nknize] for sharing the algorithm, looks pretty powerful.
I had a look into the patch and the first thing I notice is that the test
{{testLUCENE8669}} is failing. It seems that the indexed polygons are never
added to the index so fix is trivial (Method {{w.addDocument(doc)}} has been
removed in the patch):
{code:java}
- Field[] fields = LatLonShape.createIndexableFields("test", indexPoly1);
- for (Field f : fields) {
- doc.add(f);
- }
- fields = LatLonShape.createIndexableFields("test", indexPoly2);
- for (Field f : fields) {
- doc.add(f);
- }
- w.addDocument(doc);
+ addPolygonsToDoc(FIELDNAME, doc, indexPoly1);
+ addPolygonsToDoc(FIELDNAME, doc, indexPoly2);
w.forceMerge(1);{code}
Regarding the new approach, it seems there is still something missing as it
does not matter to make the methods more precise if we do not take any action
regarding the distortion of polygons due to quatization.
If we change the test by translating the polygons one degree north and 1 degree
east, the change does not have effect due to the encoding of the indexed
polygons:
{code:java}
public void testWithinFailure() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter w = new RandomIndexWriter(random(), dir);
// test polygons:
Polygon indexPoly1 = new Polygon(new double[] {4d + 1d, 4d + 1d, 3d + 1d,
3d + 1d, 4d + 1d}, new double[] {3d + 1d, 4d + 1d, 4d + 1d, 3d + 1d, 3d + 1d});
Polygon indexPoly2 = new Polygon(new double[] {2d + 1d, 2d + 1d, 1d + 1d,
1d + 1d, 2d + 1d}, new double[] {6d + 1d, 7d + 1d, 7d + 1d, 6d + 1d, 6d + 1d});
Polygon indexPoly3 = new Polygon(new double[] {1d + 1d, 1d + 1d, 0d + 1d,
0d + 1d, 1d + 1d}, new double[] {3d + 1d, 4d + 1d, 4d + 1d, 3d + 1d, 3d + 1d});
Polygon indexPoly4 = new Polygon(new double[] {2d + 1d, 2d + 1d, 1d + 1d,
1d + 1d, 2d + 1d}, new double[] {0d + 1d, 1d + 1d, 1d + 1d, 0d + 1d, 0d + 1d});
// index polygons:
Document doc;
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly1);
w.addDocument(doc);
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly2);
w.addDocument(doc);
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly3);
w.addDocument(doc);
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly4);
w.addDocument(doc);
///// search //////
IndexReader reader = w.getReader();
w.close();
IndexSearcher searcher = newSearcher(reader);
Polygon[] searchPoly = new Polygon[] {new Polygon(new double[] {4d + 1d, 4d
+ 1d, 0d + 1d, 0d + 1d, 4d + 1d}, new double[] {0d + 1d, 7d + 1d, 7d + 1d, 0d +
1d, 0d + 1d})};
Query q = LatLonShape.newPolygonQuery(FIELDNAME, QueryRelation.WITHIN,
searchPoly);
assertEquals(4, searcher.count(q));
IOUtils.close(w, reader, dir);
}{code}
I have tried to quantize the query polygon but that seems to add other issues.
was (Author: ivera):
Thanks [~nknize] for sharing the algorithm, looks pretty powerful.
I had a look into the patch and the first thing I notice is that the test
{{testLUCENE8669}} is failing. It seems that the indexed polygons are never
added to the index so fix is trivial (Method {{w.addDocument(doc)}} has been
removed in the patch):
{code:java}
- Field[] fields = LatLonShape.createIndexableFields("test", indexPoly1);
- for (Field f : fields) {
- doc.add(f);
- }
- fields = LatLonShape.createIndexableFields("test", indexPoly2);
- for (Field f : fields) {
- doc.add(f);
- }
- w.addDocument(doc);
+ addPolygonsToDoc(FIELDNAME, doc, indexPoly1);
+ addPolygonsToDoc(FIELDNAME, doc, indexPoly2);
w.forceMerge(1);{code}
Regarding the new approach, it seems there is still something missing. If we
change the test by translating the polygons one degree north and 1 degree east,
the change does not have effect due to the encoding of the indexed polygons:
{code:java}
public void testWithinFailure() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter w = new RandomIndexWriter(random(), dir);
// test polygons:
Polygon indexPoly1 = new Polygon(new double[] {4d + 1d, 4d + 1d, 3d + 1d,
3d + 1d, 4d + 1d}, new double[] {3d + 1d, 4d + 1d, 4d + 1d, 3d + 1d, 3d + 1d});
Polygon indexPoly2 = new Polygon(new double[] {2d + 1d, 2d + 1d, 1d + 1d,
1d + 1d, 2d + 1d}, new double[] {6d + 1d, 7d + 1d, 7d + 1d, 6d + 1d, 6d + 1d});
Polygon indexPoly3 = new Polygon(new double[] {1d + 1d, 1d + 1d, 0d + 1d,
0d + 1d, 1d + 1d}, new double[] {3d + 1d, 4d + 1d, 4d + 1d, 3d + 1d, 3d + 1d});
Polygon indexPoly4 = new Polygon(new double[] {2d + 1d, 2d + 1d, 1d + 1d,
1d + 1d, 2d + 1d}, new double[] {0d + 1d, 1d + 1d, 1d + 1d, 0d + 1d, 0d + 1d});
// index polygons:
Document doc;
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly1);
w.addDocument(doc);
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly2);
w.addDocument(doc);
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly3);
w.addDocument(doc);
addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly4);
w.addDocument(doc);
///// search //////
IndexReader reader = w.getReader();
w.close();
IndexSearcher searcher = newSearcher(reader);
Polygon[] searchPoly = new Polygon[] {new Polygon(new double[] {4d + 1d, 4d
+ 1d, 0d + 1d, 0d + 1d, 4d + 1d}, new double[] {0d + 1d, 7d + 1d, 7d + 1d, 0d +
1d, 0d + 1d})};
Query q = LatLonShape.newPolygonQuery(FIELDNAME, QueryRelation.WITHIN,
searchPoly);
assertEquals(4, searcher.count(q));
IOUtils.close(w, reader, dir);
}{code}
I have tried to quantize the query polygon but that seems to add other issues.
> LatLonShapePolygonQuery returning incorrect WITHIN results with shared
> boundaries
> ---------------------------------------------------------------------------------
>
> Key: LUCENE-8736
> URL: https://issues.apache.org/jira/browse/LUCENE-8736
> Project: Lucene - Core
> Issue Type: Bug
> Reporter: Nicholas Knize
> Priority: Major
> Attachments: LUCENE-8736.patch
>
>
> Triangles that are {{WITHIN}} a target polygon query that also share a
> boundary with the polygon are incorrectly reported as {{CROSSES}} instead of
> {{INSIDE}}. This leads to incorrect {{WITHIN}} query results as demonstrated
> in the following test:
> {code:java}
> public void testWithinFailure() throws Exception {
> Directory dir = newDirectory();
> RandomIndexWriter w = new RandomIndexWriter(random(), dir);
> // test polygons:
> Polygon indexPoly1 = new Polygon(new double[] {4d, 4d, 3d, 3d, 4d}, new
> double[] {3d, 4d, 4d, 3d, 3d});
> Polygon indexPoly2 = new Polygon(new double[] {2d, 2d, 1d, 1d, 2d}, new
> double[] {6d, 7d, 7d, 6d, 6d});
> Polygon indexPoly3 = new Polygon(new double[] {1d, 1d, 0d, 0d, 1d}, new
> double[] {3d, 4d, 4d, 3d, 3d});
> Polygon indexPoly4 = new Polygon(new double[] {2d, 2d, 1d, 1d, 2d}, new
> double[] {0d, 1d, 1d, 0d, 0d});
> // index polygons:
> Document doc;
> addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly1);
> w.addDocument(doc);
> addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly2);
> w.addDocument(doc);
> addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly3);
> w.addDocument(doc);
> addPolygonsToDoc(FIELDNAME, doc = new Document(), indexPoly4);
> w.addDocument(doc);
> ///// search //////
> IndexReader reader = w.getReader();
> w.close();
> IndexSearcher searcher = newSearcher(reader);
> Polygon[] searchPoly = new Polygon[] {new Polygon(new double[] {4d, 4d,
> 0d, 0d, 4d}, new double[] {0d, 7d, 7d, 0d, 0d})};
> Query q = LatLonShape.newPolygonQuery(FIELDNAME, QueryRelation.WITHIN,
> searchPoly);
> assertEquals(4, searcher.count(q));
> IOUtils.close(w, reader, dir);
> }
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
