Package elki.index.tree.spatial.kd
Class MinimalisticMemoryKDTree.KDTreePrioritySearcher
- java.lang.Object
-
- elki.index.tree.spatial.kd.MinimalisticMemoryKDTree.KDTreePrioritySearcher
-
- All Implemented Interfaces:
DBIDIter,DBIDRef,KNNSearcher<O>,PrioritySearcher<O>,RangeSearcher<O>,Iter
- Enclosing class:
- MinimalisticMemoryKDTree<O extends NumberVector>
public class MinimalisticMemoryKDTree.KDTreePrioritySearcher extends java.lang.Object implements PrioritySearcher<O>
Priority search for the k-d-tree.- Author:
- Erich Schubert
-
-
Field Summary
Fields Modifier and Type Field Description private MinimalisticMemoryKDTree.PrioritySearchBranchcurCurrent search position.private PrimitiveDistance<? super O>distanceDistance to use.private ComparableMinHeap<MinimalisticMemoryKDTree.PrioritySearchBranch>heapMin heap for searching.private DBIDArrayIteriterSearch iterator.private intposPosition within leaf.private OqueryCurrent query object.private doublethresholdStopping threshold.
-
Constructor Summary
Constructors Constructor Description KDTreePrioritySearcher(PrimitiveDistance<? super O> distance)Constructor.
-
Method Summary
All Methods Instance Methods Concrete Methods Modifier and Type Method Description PrioritySearcher<O>advance()Moves the iterator forward to the next entry.doubleallLowerBound()Lower bound for all subsequent instances (that have been completely explored).doublecomputeExactDistance()Compute the exact distance to the current candidate.PrioritySearcher<O>decreaseCutoff(double threshold)Decrease the cutoff threshold.doublegetLowerBound()Get the lower bound (if available).intinternalGetIndex()Internal only: Get the internal index.PrioritySearcher<O>search(O query)Start search with a new object.booleanvalid()Returns true if the iterator currently points to a valid object.-
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
-
Methods inherited from interface elki.database.query.PrioritySearcher
getApproximateAccuracy, getApproximateDistance, getKNN, getRange, getUpperBound, search
-
Methods inherited from interface elki.database.query.range.RangeSearcher
getRange
-
-
-
-
Field Detail
-
distance
private PrimitiveDistance<? super O extends NumberVector> distance
Distance to use.
-
heap
private ComparableMinHeap<MinimalisticMemoryKDTree.PrioritySearchBranch> heap
Min heap for searching.
-
iter
private DBIDArrayIter iter
Search iterator.
-
query
private O extends NumberVector query
Current query object.
-
threshold
private double threshold
Stopping threshold.
-
pos
private int pos
Position within leaf.
-
cur
private MinimalisticMemoryKDTree.PrioritySearchBranch cur
Current search position.
-
-
Constructor Detail
-
KDTreePrioritySearcher
public KDTreePrioritySearcher(PrimitiveDistance<? super O> distance)
Constructor.- Parameters:
distance- Distance to use
-
-
Method Detail
-
search
public PrioritySearcher<O> search(O query)
Description copied from interface:PrioritySearcherStart search with a new object.- Specified by:
searchin interfacePrioritySearcher<O extends NumberVector>- Parameters:
query- Query object- Returns:
this, for chaining
-
advance
public PrioritySearcher<O> advance()
Description copied from interface:IterMoves the iterator forward to the next entry.- Specified by:
advancein interfaceDBIDIter- Specified by:
advancein interfaceIter- Specified by:
advancein interfacePrioritySearcher<O extends NumberVector>- Returns:
- The iterator itself.
-
valid
public boolean valid()
Description copied from interface:IterReturns true if the iterator currently points to a valid object.
-
getLowerBound
public double getLowerBound()
Description copied from interface:PrioritySearcherGet the lower bound (if available).Note: the lower bound is already checked by the cutoff of the priority search, so this is primarily useful for analyzing the search behavior.
- Specified by:
getLowerBoundin interfacePrioritySearcher<O extends NumberVector>- Returns:
Double.NaNif not valid
-
allLowerBound
public double allLowerBound()
Description copied from interface:PrioritySearcherLower bound for all subsequent instances (that have been completely explored). The searcher guarantees that no further results will be returned with a distance less than this.- Specified by:
allLowerBoundin interfacePrioritySearcher<O extends NumberVector>- Returns:
- lower bound;
0if no guarantees (e.g., linear scan)
-
computeExactDistance
public double computeExactDistance()
Description copied from interface:PrioritySearcherCompute the exact distance to the current candidate.The searcher may or may not have this value already.
- Specified by:
computeExactDistancein interfacePrioritySearcher<O extends NumberVector>- Returns:
- Distance
-
internalGetIndex
public int internalGetIndex()
Description copied from interface:DBIDRefInternal only: Get the internal index.NOT FOR PUBLIC USE - ELKI Optimization engine only.
- Specified by:
internalGetIndexin interfaceDBIDRef- Returns:
- Internal index
-
decreaseCutoff
public PrioritySearcher<O> decreaseCutoff(double threshold)
Description copied from interface:PrioritySearcherDecrease the cutoff threshold.The cutoff must not be increased, as the search may have pruned some results automatically.
- Specified by:
decreaseCutoffin interfacePrioritySearcher<O extends NumberVector>- Parameters:
threshold- Threshold parameter- Returns:
- this, for chaining
-
-