public void TestNearestNeighboursEmpty()
{
var tree = new STRtree <Geometry>();
object[] nn = tree.NearestNeighbour(new GeometryItemDistance());
Assert.That(nn, Is.Null);
}
public void TestKNearestNeighbors()
{
int topK = 1000;
int totalRecords = 10000;
var geometryFactory = new GeometryFactory();
var coordinate = new Coordinate(10.1, -10.1);
var queryCenter = geometryFactory.CreatePoint(coordinate);
int valueRange = 1000;
var testDataset = new List <Geometry>();
var correctData = new List <Geometry>();
var random = new Random();
var distanceComparator = new GeometryDistanceComparer(queryCenter, true);
/*
* Generate the random test data set
*/
for (int i = 0; i < totalRecords; i++)
{
coordinate = new Coordinate(-100 + random.Next(valueRange) * 1.1, random.Next(valueRange) * (-5.1));
var spatialObject = geometryFactory.CreatePoint(coordinate);
testDataset.Add(spatialObject);
}
/*
* Sort the original data set and make sure the elements are sorted in an ascending order
*/
testDataset.Sort(distanceComparator);
/*
* Get the correct top K
*/
for (int i = 0; i < topK; i++)
{
correctData.Add(testDataset[i]);
}
var strtree = new STRtree <Geometry>();
for (int i = 0; i < totalRecords; i++)
{
strtree.Insert(testDataset[i].EnvelopeInternal, testDataset[i]);
}
/*
* Shoot a random query to make sure the STR-Tree is built.
*/
strtree.Query(new Envelope(1 + 0.1, 1 + 0.1, 2 + 0.1, 2 + 0.1));
/*
* Issue the KNN query.
*/
var testTopK = strtree.NearestNeighbour(queryCenter.EnvelopeInternal, queryCenter,
new GeometryItemDistance(), topK);
var topKList = new List <Geometry>(testTopK);
topKList.Sort(distanceComparator);
/*
* Check the difference between correct result and test result. The difference should be 0.
*/
int difference = 0;
for (int i = 0; i < topK; i++)
{
if (distanceComparator.Compare(correctData[i], topKList[i]) != 0)
{
difference++;
}
}
Assert.That(difference, Is.Zero);
}