mikera.vectorz.impl

Class SparseImmutableVector

java.lang.Object

mikera.arrayz.impl.AbstractArray<Double>

mikera.vectorz.AVector

mikera.vectorz.impl.ASizedVector

mikera.vectorz.impl.ASparseVector

mikera.vectorz.impl.ASparseIndexedVector

mikera.vectorz.impl.SparseImmutableVector

All Implemented Interfaces:

Serializable, Cloneable, Comparable<AVector>, Iterable<Double>, INDArray, ISparse, ISparseVector, IVector

public class SparseImmutableVector
extends ASparseIndexedVector

Indexed sparse immutable vector Efficient for sparse vectors. Maintains a indexed array of strictly non-zero elements. Guaranteed to contain at least one non-sparse index value

Author:

Mike

See Also:

Serialized Form

Field Summary

Fields inherited from class mikera.vectorz.impl.ASizedVector

length

Method Summary

Modifier and Type	Method and Description
void	abs() Replaces all elements of this array with their absolute values, according to Math.abs(double)
void	add(ASparseVector v)
void	addAt(int i, double value) Adds a value to a specific element of the vector This function does not perform bounds checking, i.e.
void	applyOp(Op op) Applies a unary operator to all elements of the array (in-place)
SparseIndexedVector	clone() Clones the vector, creating a new mutable copy of all data.
static AVector	create(ASparseVector source)
static AVector	create(AVector source) Creates a sparse immutable vector from the given vector, ignoring the zeros in the source.
static AVector	create(int length, Index sparseIndex, AVector sparseValues) Creates a SparseImmutableVector using the given sorted Index to identify the indexes of non-zero values, and a dense vector to specify all the non-zero element values
static AVector	create(int length, Index sparseIndex, double[] sparseValues) Creates a SparseImmutableVector using the given sorted Index to identify the indexes of non-zero values, and a double[] array to specify all the non-zero element values
static AVector	createFromRow(AMatrix m, int row) Creates a SparseIndexedVector from a row of an existing matrix
Vector	dense() Coerces this INDArray to a dense format, without changing its element values.
SparseImmutableVector	exactClone() Returns an exact clone of this vector, i.e.
double	get(int i) Returns the double value at the specified position in a 1D vector
boolean	includesIndex(int i) Returns true iff the sparse vector contains the index i
boolean	isFullyMutable() Returns true if this vector is fully mutable, i.e.
boolean	isMutable() Returns true if this vector is mutable.
boolean	isZero() Returns true if this vector is a zero vector (all components zero)
double	maxAbsElement() Returns the maximum absolute element value of a vector
int	maxAbsElementIndex() Returns the index of the maximum absolute element of a vector
int	maxElementIndex() Returns the index of the maximum element of a vector
int	minElementIndex() Returns the index of the minimum element of a vector
SparseIndexedVector	mutable() Coerces to a mutable version of a vector.
void	negate() Negates all emlements of this vector in place.
int	nonSparseElementCount() Returns the number of non-sparse elements in the sparse vector.
Index	nonSparseIndex() Returns an Index indicating which elements of this vector are defined as non-sparse.
Vector	nonSparseValues() Returns the non-sparse elements as a compacted vector
long	nonZeroCount() Returns the number of non-zero elements in the array.
int[]	nonZeroIndices() Return an int array specifying the positions in this vector which are non-zero
double[]	nonZeroValues() Return an double array specifying the values in this vector which are non-zero
void	set(AVector v) Sets the vector to equal the value of another vector
void	set(int i, double value) Sets a value at a given position in a mutable 1D array
SparseIndexedVector	sparseClone() Clones the vector into a sparse mutable format
double	unsafeGet(int i)
void	validate() Validates the internal data structure of the INDArray.
static SparseImmutableVector	wrap(int length, Index index, double[] nonSparseValues) Creates a SparseImmutableVector with the specified index and data values.

Methods inherited from class mikera.vectorz.impl.ASparseIndexedVector

addMultipleToArray, addProductToArray, addProductToArray, addToArray, addToArray, addToArray, cloneIncludingIndices, copySparseValuesTo, copyTo, dotProduct, dotProduct, dotProduct, dotProduct, elementSquaredSum, elementSum, equals, equalsArray, getElements, innerProduct, isRangeZero, multiply

Methods inherited from class mikera.vectorz.impl.ASparseVector

add, addMultiple, copyTo, copyTo, distanceL1, distanceLinf, distanceSquared, dotProduct, elementAbsPowSum, elementPowSum, elementProduct, equals, getSlices, hasUncountable, isSparse, isView, reduce, reduce, reduceSlices, roundToZero, sparse, toDoubleArray, toSparseIndexedVector

Methods inherited from class mikera.vectorz.impl.ASizedVector

checkIndex, checkLength, checkRange, checkSameLength, checkSameLength, elementCount, equalsArray, get, getShape, getShapeClone, isSameShape, length, sliceCount

Methods inherited from class mikera.vectorz.AVector

absCopy, add, add, add, add, add, add, add, add, addAt, addCopy, addCopy, addCopy, addInnerProduct, addInnerProduct, addInnerProduct, addMultiple, addMultiple, addMultiple, addMultiple, addMultiple, addMultiple, addMultiple, addMultipleCopy, addMultipleToArray, addPower, addPower, addPower, addProduct, addProduct, addProduct, addProductCopy, addProductCopy, addWeighted, angle, applyOp, applyOp, applyOp, applyOp, applyOpCopy, asDoubleArray, asElementList, asVector, broadcast, broadcastLike, broadcastLike, broadcastLike, checkDimension, checkSameLength, clamp, clampMax, clampMin, compareTo, copy, copyTo, copyTo, crossProduct, crossProduct, denseClone, dimensionality, distance, divide, divide, divide, divide, divideCopy, divideCopy, divideTo, dotProduct, elementIterator, elementMax, elementMin, elementsEqual, epsilonEquals, epsilonEquals, epsilonEquals, epsilonEquals, equals, equals, fill, fillRange, get, get, get, get, getElements, getElements, getLongShape, getShape, getTranspose, getTransposeCopy, getTransposeView, hashCode, immutable, innerProduct, innerProduct, innerProduct, innerProduct, innerProduct, interpolate, isElementConstrained, isSameShape, isUnitLengthVector, isUnitLengthVector, iterator, join, join, join, log, logistic, magnitude, magnitudeSquared, maxElement, minElement, multiply, multiply, multiply, multiply, multiplyCopy, multiplyCopy, multiplyCopy, multiplyTo, negateCopy, normalise, normaliseCopy, normaliseMaxAbsElement, outerProduct, outerProduct, pow, projectToPlane, reduceSlices, reorder, reorder, reshape, rotateCopy, rotateView, rotateView, scale, scaleAdd, scaleAdd, scaleAdd, scaleAdd, scaleCopy, scaleToMagnitude, select, selectClone, selectView, set, set, set, set, set, set, set, set, set, setElements, setElements, setElements, setInnerProduct, setInnerProduct, setInnerProduct, setMultiple, setMultiple, setMultiple, setMultiple, shiftCopy, signum, slice, slice, sliceValue, softmax, softmaxCopy, sqrtCopy, square, squareCopy, sub, sub, subArray, subAt, subCopy, subCopy, subMultiple, subVector, tanh, toDoubleBuffer, toList, toNormal, toSliceArray, toString, toStringFull, toVector, tryEfficientJoin, unsafeSet

Methods inherited from class mikera.arrayz.impl.AbstractArray

add, addInnerProduct, addInnerProduct, addOuterProduct, broadcastCloneLike, broadcastCopyLike, checkMutable, componentCount, density, ensureMutable, exp, get, get, get, getComponent, getComponents, getElements, getElements, getSlices, getSliceViews, isBoolean, isDense, reciprocal, reciprocalCopy, scale, set, set, signumCopy, sqrt, sub, sub, toArray, withComponents

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, wait, wait, wait

Methods inherited from interface mikera.vectorz.IVector

addCopy, divideCopy, epsilonEquals, isSameShape, length, multiplyCopy, normalise, normaliseCopy, rotateCopy, rotateView, shiftCopy, sqrtCopy, subCopy

Methods inherited from interface mikera.arrayz.INDArray

absCopy, add, add, addAt, addCopy, addCopy, addInnerProduct, addMultiple, addOuterProduct, addPower, addPower, applyOp, applyOp, applyOp, applyOpCopy, asDoubleArray, asElementList, asVector, broadcast, broadcastCloneLike, broadcastCopyLike, broadcastLike, broadcastLike, broadcastLike, checkMutable, clamp, componentCount, copy, denseClone, dimensionality, divide, divide, divideCopy, elementCount, elementIterator, elementMax, elementMin, elementsEqual, ensureMutable, epsilonEquals, epsilonEquals, equals, equalsArray, exp, fill, get, get, get, get, get, get, get, get, getComponent, getComponents, getElements, getElements, getElements, getLongShape, getShape, getShape, getShapeClone, getSlices, getSliceViews, getTranspose, getTransposeCopy, getTransposeView, immutable, innerProduct, innerProduct, innerProduct, innerProduct, isBoolean, isDense, isElementConstrained, isSameShape, join, join, log, multiply, multiplyCopy, multiplyCopy, negateCopy, outerProduct, pow, reciprocal, reciprocalCopy, reduceSlices, reorder, reorder, reshape, rotateView, scale, scaleAdd, scaleAdd, scaleCopy, set, set, set, set, set, set, setElements, setElements, setElements, setInnerProduct, setMultiple, setMultiple, signum, signumCopy, slice, slice, sliceCount, sliceValue, sqrt, square, squareCopy, sub, sub, subArray, subCopy, toArray, toDoubleBuffer, toSliceArray, toVector, withComponents

Methods inherited from interface mikera.arrayz.ISparse

density

Methods inherited from interface java.lang.Iterable

forEach, spliterator

Method Detail

wrap

public static SparseImmutableVector wrap(int length,
                                         Index index,
                                         double[] nonSparseValues)

Creates a SparseImmutableVector with the specified index and data values. WARNING: Performs no checking - Index must be distinct and sorted, and data must be non-zero.

create

public static AVector create(int length,
                             Index sparseIndex,
                             double[] sparseValues)

Creates a SparseImmutableVector using the given sorted Index to identify the indexes of non-zero values, and a double[] array to specify all the non-zero element values

create

public static AVector create(int length,
                             Index sparseIndex,
                             AVector sparseValues)

Creates a SparseImmutableVector using the given sorted Index to identify the indexes of non-zero values, and a dense vector to specify all the non-zero element values

create

public static AVector create(AVector source)

Creates a sparse immutable vector from the given vector, ignoring the zeros in the source.

create

public static AVector create(ASparseVector source)

createFromRow

public static AVector createFromRow(AMatrix m,
                                    int row)

Creates a SparseIndexedVector from a row of an existing matrix

nonSparseElementCount

public int nonSparseElementCount()

Description copied from class: ASparseVector

Returns the number of non-sparse elements in the sparse vector. This will be greater or equal to the number of non-zero elements

Specified by:

nonSparseElementCount in class ASparseVector

Returns:

isZero

public boolean isZero()

Description copied from class: AVector

Returns true if this vector is a zero vector (all components zero)

Specified by:

isZero in interface INDArray

Overrides:

isZero in class ASparseIndexedVector

maxAbsElement

public double maxAbsElement()

Description copied from class: AVector

Returns the maximum absolute element value of a vector

Overrides:

maxAbsElement in class AVector

Returns:

maxElementIndex

public int maxElementIndex()

Description copied from class: AVector

Returns the index of the maximum element of a vector

Overrides:

maxElementIndex in class AVector

Returns:

maxAbsElementIndex

public int maxAbsElementIndex()

Description copied from class: AVector

Returns the index of the maximum absolute element of a vector

Overrides:

maxAbsElementIndex in class AVector

Returns:

minElementIndex

public int minElementIndex()

Description copied from class: AVector

Returns the index of the minimum element of a vector

Overrides:

minElementIndex in class AVector

negate

public void negate()

Description copied from class: AVector

Negates all emlements of this vector in place.

Specified by:

negate in interface INDArray

Overrides:

negate in class AVector

applyOp

public void applyOp(Op op)

Description copied from interface: INDArray

Applies a unary operator to all elements of the array (in-place)

Specified by:

applyOp in interface INDArray

Overrides:

applyOp in class AVector

abs

public void abs()

Description copied from interface: INDArray

Replaces all elements of this array with their absolute values, according to Math.abs(double)

Specified by:

abs in interface INDArray

Overrides:

abs in class AVector

get

public double get(int i)

Description copied from interface: INDArray

Returns the double value at the specified position in a 1D vector

Specified by:

get in interface INDArray

Specified by:

get in interface IVector

Specified by:

get in class AVector

unsafeGet

public double unsafeGet(int i)

Overrides:

unsafeGet in class AVector

isFullyMutable

public boolean isFullyMutable()

Description copied from class: AVector

Returns true if this vector is fully mutable, i.e. can contain any unconstrained double values

Specified by:

isFullyMutable in interface INDArray

Overrides:

isFullyMutable in class AVector

isMutable

public boolean isMutable()

Description copied from class: AVector

Returns true if this vector is mutable.

Specified by:

isMutable in interface INDArray

Overrides:

isMutable in class AVector

nonZeroCount

public long nonZeroCount()

Description copied from interface: INDArray

Returns the number of non-zero elements in the array.

Specified by:

nonZeroCount in interface INDArray

Overrides:

nonZeroCount in class ASparseIndexedVector

nonZeroIndices

public int[] nonZeroIndices()

Description copied from class: AVector

Return an int array specifying the positions in this vector which are non-zero

Overrides:

nonZeroIndices in class ASparseIndexedVector

Returns:

add

public void add(ASparseVector v)

Specified by:

add in class ASparseVector

set

public void set(AVector v)

Description copied from class: AVector

Sets the vector to equal the value of another vector

Overrides:

set in class AVector

set

public void set(int i,
                double value)

Description copied from interface: INDArray

Sets a value at a given position in a mutable 1D array

Specified by:

set in interface INDArray

Specified by:

set in interface IVector

Specified by:

set in class AVector

addAt

public void addAt(int i,
                  double value)

Description copied from class: AVector

Adds a value to a specific element of the vector This function does not perform bounds checking, i.e. is an unsafe operation

Overrides:

addAt in class AVector

nonSparseValues

public Vector nonSparseValues()

Description copied from interface: ISparseVector

Returns the non-sparse elements as a compacted vector

Specified by:

nonSparseValues in interface ISparseVector

Specified by:

nonSparseValues in class ASparseVector

Returns:

nonZeroValues

public double[] nonZeroValues()

Description copied from class: AVector

Return an double array specifying the values in this vector which are non-zero

Overrides:

nonZeroValues in class AVector

Returns:

nonSparseIndex

public Index nonSparseIndex()

Description copied from class: AVector

Returns an Index indicating which elements of this vector are defined as non-sparse. A sparse index must be zero. A non-sparse element may be zero, but is expected be non-zero.

Specified by:

nonSparseIndex in interface ISparseVector

Specified by:

nonSparseIndex in class ASparseVector

Returns:

includesIndex

public boolean includesIndex(int i)

Description copied from class: ASparseVector

Returns true iff the sparse vector contains the index i

Overrides:

includesIndex in class ASparseIndexedVector

Returns:

dense

public Vector dense()

Description copied from interface: INDArray

Coerces this INDArray to a dense format, without changing its element values. May return the same INDArray if already dense. May also mutate the internal structure of the original NDArray, or create a view over parts of the original INDArray. You should take a defensive copy of the original NDArray if any of this concerns you. The returned dense array may not be fully mutable in all elements.

Specified by:

dense in interface INDArray

Overrides:

dense in class AVector

mutable

public SparseIndexedVector mutable()

Description copied from class: AVector

Coerces to a mutable version of a vector. May or may not be a copy, but guaranteed to be fully mutable

Specified by:

mutable in interface INDArray

Overrides:

mutable in class AVector

clone

public SparseIndexedVector clone()

Description copied from class: AVector

Clones the vector, creating a new mutable copy of all data. The clone is: - not guaranteed to be of the same type. - guaranteed to be fully mutable - guaranteed not to contain a reference (i.e. is a full deep copy)

Specified by:

clone in interface INDArray

Specified by:

clone in interface IVector

Overrides:

clone in class ASparseVector

sparseClone

public SparseIndexedVector sparseClone()

Description copied from class: AVector

Clones the vector into a sparse mutable format

Specified by:

sparseClone in interface INDArray

Overrides:

sparseClone in class AVector

exactClone

public SparseImmutableVector exactClone()

Description copied from class: AVector

Returns an exact clone of this vector, i.e. of the same type

Specified by:

exactClone in interface INDArray

Specified by:

exactClone in class AVector

Returns:

A clone of the original array

validate

public void validate()

Description copied from interface: INDArray

Validates the internal data structure of the INDArray. Throws an exception on failure. Failure indicates a serious bug and/or data corruption.

Specified by:

validate in interface INDArray

Overrides:

validate in class AVector