NV
- the type of NumberVector handled by this AlgorithmD
- the type of Distance used@Title(value="DeliClu: Density-Based Hierarchical Clustering") @Description(value="Hierachical algorithm to find density-connected sets in a database based on the parameter \'minpts\'.") @Reference(authors="E. Achtert, C. B\u00f6hm, P. Kr\u00f6ger", title="DeLiClu: Boosting Robustness, Completeness, Usability, and Efficiency of Hierarchical Clustering by a Closest Pair Ranking", booktitle="Proc. 10th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD 2006), Singapore, 2006", url="http://dx.doi.org/10.1007/11731139_16") public class DeLiClu<NV extends NumberVector<?>,D extends Distance<D>> extends AbstractDistanceBasedAlgorithm<NV,D,ClusterOrderResult<D>> implements OPTICSTypeAlgorithm<D>
Reference:
E. Achtert, C. Böhm, P. Kröger: DeLiClu: Boosting Robustness, Completeness,
Usability, and Efficiency of Hierarchical Clustering by a Closest Pair
Ranking.
In Proc. 10th Pacific-Asia Conference on Knowledge Discovery and Data Mining
(PAKDD 2006), Singapore, 2006.
Modifier and Type | Class and Description |
---|---|
static class |
DeLiClu.Parameterizer<NV extends NumberVector<?>,D extends Distance<D>>
Parameterization class.
|
class |
DeLiClu.SpatialObjectPair
Encapsulates an entry in the cluster order.
|
Modifier and Type | Field and Description |
---|---|
private UpdatableHeap<DeLiClu.SpatialObjectPair> |
heap
The priority queue for the algorithm.
|
private KNNJoin<NV,D,DeLiCluNode,DeLiCluEntry> |
knnJoin
Holds the knnJoin algorithm.
|
private static Logging |
LOG
The logger for this class.
|
private int |
minpts
Holds the value of
MINPTS_ID . |
static OptionID |
MINPTS_ID
Parameter to specify the threshold for minimum number of points within a
cluster, must be an integer greater than 0.
|
DISTANCE_FUNCTION_ID
Constructor and Description |
---|
DeLiClu(DistanceFunction<? super NV,D> distanceFunction,
int minpts)
Constructor.
|
Modifier and Type | Method and Description |
---|---|
private void |
expandDirNodes(SpatialPrimitiveDistanceFunction<NV,D> distFunction,
DeLiCluNode node1,
DeLiCluNode node2)
Expands the specified directory nodes.
|
private void |
expandLeafNodes(SpatialPrimitiveDistanceFunction<NV,D> distFunction,
DeLiCluNode node1,
DeLiCluNode node2,
DataStore<KNNList<D>> knns)
Expands the specified leaf nodes.
|
private void |
expandNodes(DeLiCluTree index,
SpatialPrimitiveDistanceFunction<NV,D> distFunction,
DeLiClu.SpatialObjectPair nodePair,
DataStore<KNNList<D>> knns)
Expands the spatial nodes of the specified pair.
|
D |
getDistanceFactory()
Get the distance factory.
|
TypeInformation[] |
getInputTypeRestriction()
Get the input type restriction used for negotiating the data query.
|
protected Logging |
getLogger()
Get the (STATIC) logger for this class.
|
int |
getMinPts()
Get the minpts value used.
|
private DBID |
getStartObject(Relation<NV> relation)
Returns the id of the start object for the run method.
|
private void |
reinsertExpanded(SpatialPrimitiveDistanceFunction<NV,D> distFunction,
DeLiCluTree index,
List<TreeIndexPathComponent<DeLiCluEntry>> path,
DataStore<KNNList<D>> knns)
Reinserts the objects of the already expanded nodes.
|
private void |
reinsertExpanded(SpatialPrimitiveDistanceFunction<NV,D> distFunction,
DeLiCluTree index,
List<TreeIndexPathComponent<DeLiCluEntry>> path,
int pos,
SpatialDirectoryEntry parentEntry,
DataStore<KNNList<D>> knns) |
ClusterOrderResult<D> |
run(Database database,
Relation<NV> relation) |
getDistanceFunction
makeParameterDistanceFunction, run
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
run
private static final Logging LOG
public static final OptionID MINPTS_ID
private UpdatableHeap<DeLiClu.SpatialObjectPair> heap
private KNNJoin<NV extends NumberVector<?>,D extends Distance<D>,DeLiCluNode,DeLiCluEntry> knnJoin
private int minpts
MINPTS_ID
.public DeLiClu(DistanceFunction<? super NV,D> distanceFunction, int minpts)
distanceFunction
- Distance functionminpts
- MinPtspublic ClusterOrderResult<D> run(Database database, Relation<NV> relation)
private DBID getStartObject(Relation<NV> relation)
relation
- the database relation storing the objectsprivate void expandNodes(DeLiCluTree index, SpatialPrimitiveDistanceFunction<NV,D> distFunction, DeLiClu.SpatialObjectPair nodePair, DataStore<KNNList<D>> knns)
index
- the index storing the objectsdistFunction
- the spatial distance function of this algorithmnodePair
- the pair of nodes to be expandedknns
- the knn listprivate void expandDirNodes(SpatialPrimitiveDistanceFunction<NV,D> distFunction, DeLiCluNode node1, DeLiCluNode node2)
distFunction
- the spatial distance function of this algorithmnode1
- the first nodenode2
- the second nodeprivate void expandLeafNodes(SpatialPrimitiveDistanceFunction<NV,D> distFunction, DeLiCluNode node1, DeLiCluNode node2, DataStore<KNNList<D>> knns)
distFunction
- the spatial distance function of this algorithmnode1
- the first nodenode2
- the second nodeknns
- the knn listprivate void reinsertExpanded(SpatialPrimitiveDistanceFunction<NV,D> distFunction, DeLiCluTree index, List<TreeIndexPathComponent<DeLiCluEntry>> path, DataStore<KNNList<D>> knns)
distFunction
- the spatial distance function of this algorithmindex
- the index storing the objectspath
- the path of the object inserted lastknns
- the knn listprivate void reinsertExpanded(SpatialPrimitiveDistanceFunction<NV,D> distFunction, DeLiCluTree index, List<TreeIndexPathComponent<DeLiCluEntry>> path, int pos, SpatialDirectoryEntry parentEntry, DataStore<KNNList<D>> knns)
public int getMinPts()
OPTICSTypeAlgorithm
getMinPts
in interface OPTICSTypeAlgorithm<D extends Distance<D>>
public D getDistanceFactory()
OPTICSTypeAlgorithm
getDistanceFactory
in interface OPTICSTypeAlgorithm<D extends Distance<D>>
public TypeInformation[] getInputTypeRestriction()
AbstractAlgorithm
getInputTypeRestriction
in interface Algorithm
getInputTypeRestriction
in class AbstractAlgorithm<ClusterOrderResult<D extends Distance<D>>>
protected Logging getLogger()
AbstractAlgorithm
getLogger
in class AbstractAlgorithm<ClusterOrderResult<D extends Distance<D>>>