O
- The object typeD
- The distance typepublic static class PartitionApproximationMaterializeKNNPreprocessor.Factory<O,D extends Distance<D>> extends AbstractMaterializeKNNPreprocessor.Factory<O,D,KNNList<D>>
Modifier and Type | Class and Description |
---|---|
static class |
PartitionApproximationMaterializeKNNPreprocessor.Factory.Parameterizer<O,D extends Distance<D>>
Parameterization class.
|
Modifier and Type | Field and Description |
---|---|
(package private) int |
partitions
The number of partitions to use
|
private RandomFactory |
rnd
Random generator
|
DISTANCE_FUNCTION_ID, distanceFunction, k, K_ID
Constructor and Description |
---|
PartitionApproximationMaterializeKNNPreprocessor.Factory(int k,
DistanceFunction<? super O,D> distanceFunction,
int partitions,
RandomFactory rnd)
Constructor.
|
Modifier and Type | Method and Description |
---|---|
PartitionApproximationMaterializeKNNPreprocessor<O,D> |
instantiate(Relation<O> relation)
Sets the database in the distance function of this index (if existing).
|
getDistanceFactory, getDistanceFunction, getInputTypeRestriction
int partitions
private final RandomFactory rnd
public PartitionApproximationMaterializeKNNPreprocessor.Factory(int k, DistanceFunction<? super O,D> distanceFunction, int partitions, RandomFactory rnd)
k
- kdistanceFunction
- distance functionpartitions
- number of partitionsrnd
- public PartitionApproximationMaterializeKNNPreprocessor<O,D> instantiate(Relation<O> relation)
IndexFactory
instantiate
in interface IndexFactory<O,KNNIndex<O>>
instantiate
in class AbstractMaterializeKNNPreprocessor.Factory<O,D extends Distance<D>,KNNList<D extends Distance<D>>>
relation
- the relation to index