org.nd4j.linalg.api.ops

Class BaseAccumulation

java.lang.Object

org.nd4j.autodiff.functions.DifferentialFunction

org.nd4j.linalg.api.ops.BaseOp

org.nd4j.linalg.api.ops.BaseAccumulation

All Implemented Interfaces:

Accumulation, Op

Direct Known Subclasses:

All, AMax, AMin, Any, ASum, Bias, CosineDistance, CosineSimilarity, CountNonZero, CountZero, Dot, Entropy, EqualsWithEps, EuclideanDistance, HammingDistance, JaccardDistance, LogEntropy, LogSumExp, ManhattanDistance, MatchCondition, Max, Min, Norm1, Norm2, NormMax, Prod, ShannonEntropy, Sum, Variance

public abstract class BaseAccumulation
extends BaseOp
implements Accumulation

Base class for accumulation, initiates the initial entry with respect to the child class. Also contains baseline fields for the over all field with accumulation.

Author:

Adam Gibson

Nested Class Summary

Nested classes/interfaces inherited from interface org.nd4j.linalg.api.ops.Op

Op.Type

Field Summary

Fields

Modifier and Type	Field and Description
protected Number	finalResult
protected boolean	isComplex
protected boolean	keepDims
protected boolean	newFormat

Fields inherited from class org.nd4j.linalg.api.ops.BaseOp

extraArgs, extraArgz, n, numProcessed, passThrough, x, xVertexId, y, yVertexId, z, zVertexId

Fields inherited from class org.nd4j.autodiff.functions.DifferentialFunction

dimensions, inPlace, sameDiff, scalarValue

Constructor Summary

Constructors

Constructor and Description
BaseAccumulation()
BaseAccumulation(INDArray x)
BaseAccumulation(INDArray x, INDArray y)
BaseAccumulation(INDArray x, INDArray y, INDArray z, long n) Initialize with the given input, pairwise transform, result, and number of elements
BaseAccumulation(INDArray x, INDArray y, long n)
BaseAccumulation(SameDiff sameDiff)
BaseAccumulation(SameDiff sameDiff, SDVariable i_v)
BaseAccumulation(SameDiff sameDiff, SDVariable i_v, int[] dimensions)
BaseAccumulation(SameDiff sameDiff, SDVariable i_v, int[] dimensions, boolean keepDims)
BaseAccumulation(SameDiff sameDiff, SDVariable i_v, SDVariable i_v2, int[] dimensions)
BaseAccumulation(SameDiff sameDiff, SDVariable i_v, SDVariable i_v2, int[] dimensions, boolean keepDims)

Method Summary

Modifier and Type	Method and Description
List<int[]>	calculateOutputShape() Calculate the output shape for this op
Number	getFinalResult() Get the final result (may return null if getAndSetFinalResult has not been called, or for accumulation ops on complex arrays)
protected boolean	hasReductionIndices(NodeDef nodeDef)
void	initFromOnnx(OnnxProto3.NodeProto node, SameDiff initWith, Map<String,OnnxProto3.AttributeProto> attributesForNode, OnnxProto3.GraphProto graph) Iniitialize the function from the given OnnxProto3.NodeProto
void	initFromTensorFlow(NodeDef nodeDef, SameDiff initWith, Map<String,AttrValue> attributesForNode, GraphDef graph) Initialize the function from the given NodeDef
boolean	isComplexAccumulation()
boolean	isKeepDims() This method returns TRUE if we're going to keep axis, FALSE otherwise
INDArray	noOp() Returns the no op version of the input Basically when a reduce can't happen (eg: sum(0) on a row vector) you have a no op state for a given reduction.
Op.Type	opType() The type of the op
void	setFinalResult(double value) Get the final result (may return null if getAndSetFinalResult has not been called, or for accumulation ops on complex arrays)
double	zeroDouble() Initial value (used to initialize the accumulation op)
float	zeroFloat() Initial value (used to initialize the accumulation op)
float	zeroHalf() Initial value for half

Methods inherited from class org.nd4j.linalg.api.ops.BaseOp

equals, exec, exec, extraArgs, extraArgsBuff, extraArgsDataBuff, getOpType, hashCode, init, isExecSpecial, isPassThrough, n, numProcessed, outputVariables, setN, setX, setY, setZ, toCustomOp, toString, x, y, z

Methods inherited from class org.nd4j.autodiff.functions.DifferentialFunction

arg, args, asProperties, attributeAdaptersForFunction, configFieldName, diff, doDiff, dup, f, getValue, hasPlaceHolderInputs, isConfigProperties, larg, mappingsForFunction, onnxName, onnxNames, opName, opNum, outputVariables, propertiesForFunction, rarg, resolvePropertiesFromSameDiffBeforeExecution, setInstanceId, setValueFor, tensorflowName, tensorflowNames

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.nd4j.linalg.api.ops.Accumulation

getOpType

Methods inherited from interface org.nd4j.linalg.api.ops.Op

exec, exec, extraArgs, extraArgsBuff, extraArgsDataBuff, init, isExecSpecial, isPassThrough, n, numProcessed, opName, opNum, setExtraArgs, setN, setX, setY, setZ, toCustomOp, x, y, z

Field Detail

finalResult

protected Number finalResult

keepDims

protected boolean keepDims

newFormat

protected boolean newFormat

isComplex

protected boolean isComplex

Constructor Detail

BaseAccumulation

public BaseAccumulation(SameDiff sameDiff,
                        SDVariable i_v,
                        int[] dimensions,
                        boolean keepDims)

BaseAccumulation

public BaseAccumulation(SameDiff sameDiff,
                        SDVariable i_v,
                        SDVariable i_v2,
                        int[] dimensions,
                        boolean keepDims)

BaseAccumulation

public BaseAccumulation(SameDiff sameDiff,
                        SDVariable i_v)

BaseAccumulation

public BaseAccumulation(SameDiff sameDiff,
                        SDVariable i_v,
                        int[] dimensions)

BaseAccumulation

public BaseAccumulation(SameDiff sameDiff,
                        SDVariable i_v,
                        SDVariable i_v2,
                        int[] dimensions)

BaseAccumulation

public BaseAccumulation()

BaseAccumulation

public BaseAccumulation(INDArray x,
                        INDArray y,
                        INDArray z,
                        long n)

Initialize with the given input, pairwise transform, result, and number of elements

Parameters:

x - the input

y - the pairwise transform

z - the result

n - the number of elements

BaseAccumulation

public BaseAccumulation(INDArray x,
                        INDArray y,
                        long n)

BaseAccumulation

public BaseAccumulation(INDArray x)

BaseAccumulation

public BaseAccumulation(INDArray x,
                        INDArray y)

BaseAccumulation

public BaseAccumulation(SameDiff sameDiff)

Method Detail

noOp

public INDArray noOp()

Description copied from interface: Accumulation

Returns the no op version of the input Basically when a reduce can't happen (eg: sum(0) on a row vector) you have a no op state for a given reduction. For most accumulations, this should return x but certain transformations should return say: the absolute value

Specified by:

noOp in interface Accumulation

Returns:

the no op version of the input

isKeepDims

public boolean isKeepDims()

Description copied from interface: Accumulation

This method returns TRUE if we're going to keep axis, FALSE otherwise

Specified by:

isKeepDims in interface Accumulation

Returns:

calculateOutputShape

public List<int[]> calculateOutputShape()

Description copied from class: DifferentialFunction

Calculate the output shape for this op

Overrides:

calculateOutputShape in class DifferentialFunction

Returns:

initFromTensorFlow

public void initFromTensorFlow(NodeDef nodeDef,
                               SameDiff initWith,
                               Map<String,AttrValue> attributesForNode,
                               GraphDef graph)

Description copied from class: DifferentialFunction

Initialize the function from the given NodeDef

Overrides:

initFromTensorFlow in class BaseOp

hasReductionIndices

protected boolean hasReductionIndices(NodeDef nodeDef)

initFromOnnx

public void initFromOnnx(OnnxProto3.NodeProto node,
                         SameDiff initWith,
                         Map<String,OnnxProto3.AttributeProto> attributesForNode,
                         OnnxProto3.GraphProto graph)

Description copied from class: DifferentialFunction

Iniitialize the function from the given OnnxProto3.NodeProto

Overrides:

initFromOnnx in class BaseOp

setFinalResult

public void setFinalResult(double value)

Description copied from interface: Accumulation

Get the final result (may return null if getAndSetFinalResult has not been called, or for accumulation ops on complex arrays)

Specified by:

setFinalResult in interface Accumulation

getFinalResult

public Number getFinalResult()

Description copied from interface: Accumulation

Get the final result (may return null if getAndSetFinalResult has not been called, or for accumulation ops on complex arrays)

Specified by:

getFinalResult in interface Accumulation

zeroDouble

public double zeroDouble()

Description copied from interface: Accumulation

Initial value (used to initialize the accumulation op)

Specified by:

zeroDouble in interface Accumulation

Returns:

the initial value

zeroFloat

public float zeroFloat()

Description copied from interface: Accumulation

Initial value (used to initialize the accumulation op)

Specified by:

zeroFloat in interface Accumulation

zeroHalf

public float zeroHalf()

Description copied from interface: Accumulation

Initial value for half

Specified by:

zeroHalf in interface Accumulation

Returns:

opType

public Op.Type opType()

Description copied from class: DifferentialFunction

The type of the op

Overrides:

opType in class DifferentialFunction

Returns:

isComplexAccumulation

public boolean isComplexAccumulation()

Specified by:

isComplexAccumulation in interface Accumulation