org.tribuo.math.util

Class HeapMerger

java.lang.Object

org.tribuo.math.util.HeapMerger

All Implemented Interfaces:

Serializable, Merger

public class HeapMerger
extends Object
implements Merger

Merges each SparseVector separately using a PriorityQueue as a heap.

Relies upon VectorIterator.compareTo(VectorIterator).

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
HeapMerger()

Method Summary

Modifier and Type	Method and Description
DenseSparseMatrix	merge(DenseSparseMatrix[] inputs) Merges an array of DenseSparseMatrix into a single DenseSparseMatrix.
static SparseVector	merge(List<SparseVector> vectors, int dimension, int[] indicesBuffer, double[] valuesBuffer) Merges a list of sparse vectors into a single sparse vector, summing the values.
SparseVector	merge(SparseVector[] inputs) Merges an array of SparseVector into a single SparseVector.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

HeapMerger

public HeapMerger()

Method Detail

merge

public DenseSparseMatrix merge(DenseSparseMatrix[] inputs)

Description copied from interface: Merger

Merges an array of DenseSparseMatrix into a single DenseSparseMatrix.

Specified by:

merge in interface Merger

Parameters:

inputs - The matrices to merge.

Returns:

The merged matrix.

merge

public SparseVector merge(SparseVector[] inputs)

Description copied from interface: Merger

Merges an array of SparseVector into a single SparseVector.

Specified by:

merge in interface Merger

Parameters:

inputs - The vectors to merge.

Returns:

The merged vector.

merge

public static SparseVector merge(List<SparseVector> vectors,
                                 int dimension,
                                 int[] indicesBuffer,
                                 double[] valuesBuffer)

Merges a list of sparse vectors into a single sparse vector, summing the values.

Parameters:

vectors - The vectors to merge.

dimension - The dimension of the sparse vector.

indicesBuffer - A buffer for the indices.

valuesBuffer - A buffer for the values.

Returns:

The merged SparseVector.