org.nd4j.linalg.api.ops

Class BaseOp

java.lang.Object

org.nd4j.autodiff.functions.DifferentialFunction

org.nd4j.linalg.api.ops.BaseOp

All Implemented Interfaces:

Op

Direct Known Subclasses:

BaseAccumulation, BaseBroadcastOp, BaseGridOp, BaseIndexAccumulation, BaseRandomOp, BaseScalarOp, BaseTransformOp, DefaultOpConverter, ShapeOp

public abstract class BaseOp
extends DifferentialFunction
implements Op

Base op. An op involves iterating over 2 buffers (x,y) up to n elements and applying a transform or accumulating a result.

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 Object[]	extraArgs
protected DataBuffer	extraArgz
protected long	n
protected long	numProcessed
protected boolean	passThrough
protected INDArray	x
protected String	xVertexId
protected INDArray	y
protected String	yVertexId
protected INDArray	z
protected String	zVertexId

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

dimensions, inPlace, sameDiff, scalarValue

Constructor Summary

Constructors

Constructor and Description
BaseOp()
BaseOp(INDArray x) An op for one ndarray
BaseOp(INDArray x, INDArray z) Specify an alternative result array
BaseOp(INDArray x, INDArray y, INDArray z, long n)
BaseOp(INDArray x, INDArray z, long n) Specify an alternative output array
BaseOp(SameDiff sameDiff, boolean inPlace, Object[] extraArgs)
BaseOp(SameDiff sameDiff, Object[] extraArgs)

Method Summary

Modifier and Type	Method and Description
boolean	equals(Object o)
void	exec() Execute the op if its pass through (not needed most of the time)
void	exec(int... dimensions) Exec along each dimension
Object[]	extraArgs() Extra arguments
Buffer	extraArgsBuff() Returns a buffer of either float or double of the extra args for this buffer
DataBuffer	extraArgsDataBuff() Returns the extra args as a data buffer
static Op.Type	getOpType(Op op)
int	hashCode()
void	init(INDArray x, INDArray y, INDArray z, long n) Initialize the operation based on the parameters
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	isExecSpecial() Whether the executioner needs to do a special call or not
boolean	isPassThrough() Returns whether the op should be executed or not (through the executioner)
long	n() The number of elements to do a op over
long	numProcessed() Number processed
SDVariable[]	outputVariables(String baseName) Return the output functions for this differential function.
void	setN(long n) Change n
void	setX(INDArray x) set x (the input ndarray)
void	setY(INDArray y) set y(the pairwise ndarray)
void	setZ(INDArray z) set z (the solution ndarray)
CustomOp	toCustomOp() Converts this op to be a CustomOp A CustomOp is a more flexible op meant for multiple inputs and outputs.
String	toString()
INDArray	x() The origin ndarray
INDArray	y() The pairwise op ndarray
INDArray	z() The resulting ndarray

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

arg, args, asProperties, attributeAdaptersForFunction, calculateOutputShape, configFieldName, diff, doDiff, dup, f, getValue, hasPlaceHolderInputs, isConfigProperties, larg, mappingsForFunction, onnxName, onnxNames, opName, opNum, opType, 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.Op

opName, opNum, setExtraArgs

Field Detail

x

protected INDArray x

y

protected INDArray y

z

protected INDArray z

n

protected long n

numProcessed

protected long numProcessed

extraArgs

protected Object[] extraArgs

passThrough

protected boolean passThrough

xVertexId

protected String xVertexId

yVertexId

protected String yVertexId

zVertexId

protected String zVertexId

extraArgz

protected DataBuffer extraArgz

Constructor Detail

BaseOp

public BaseOp()

BaseOp

public BaseOp(SameDiff sameDiff,
              boolean inPlace,
              Object[] extraArgs)

BaseOp

public BaseOp(SameDiff sameDiff,
              Object[] extraArgs)

BaseOp

public BaseOp(INDArray x,
              INDArray z)

Specify an alternative result array

Parameters:

x - the input

z - the output array

BaseOp

public BaseOp(INDArray x,
              INDArray z,
              long n)

Specify an alternative output array

Parameters:

x - the input

z - the output

n - the number of elements to iterate on

BaseOp

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

BaseOp

public BaseOp(INDArray x)

An op for one ndarray

Parameters:

x - the ndarray

Method Detail

isExecSpecial

public boolean isExecSpecial()

Description copied from interface: Op

Whether the executioner needs to do a special call or not

Specified by:

isExecSpecial in interface Op

Returns:

true if the executioner needs to do a special call or not false otherwise

getOpType

public static Op.Type getOpType(Op op)

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

Specified by:

initFromTensorFlow in class DifferentialFunction

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

Specified by:

initFromOnnx in class DifferentialFunction

extraArgsDataBuff

public DataBuffer extraArgsDataBuff()

Description copied from interface: Op

Returns the extra args as a data buffer

Specified by:

extraArgsDataBuff in interface Op

Returns:

extraArgsBuff

public Buffer extraArgsBuff()

Description copied from interface: Op

Returns a buffer of either float or double of the extra args for this buffer

Specified by:

extraArgsBuff in interface Op

Returns:

a buffer of either opType float or double representing the extra args for this op

isPassThrough

public boolean isPassThrough()

Description copied from interface: Op

Returns whether the op should be executed or not (through the executioner)

Specified by:

isPassThrough in interface Op

Returns:

true if the op is pass through false otherwise

setX

public void setX(INDArray x)

Description copied from interface: Op

set x (the input ndarray)

Specified by:

setX in interface Op

setZ

public void setZ(INDArray z)

Description copied from interface: Op

set z (the solution ndarray)

Specified by:

setZ in interface Op

setY

public void setY(INDArray y)

Description copied from interface: Op

set y(the pairwise ndarray)

Specified by:

setY in interface Op

extraArgs

public Object[] extraArgs()

Description copied from interface: Op

Extra arguments

Specified by:

extraArgs in interface Op

Returns:

the extra arguments

x

public INDArray x()

Description copied from interface: Op

The origin ndarray

Specified by:

x in interface Op

Returns:

the origin ndarray

y

public INDArray y()

Description copied from interface: Op

The pairwise op ndarray

Specified by:

y in interface Op

Returns:

the pairwise op ndarray

z

public INDArray z()

Description copied from interface: Op

The resulting ndarray

Specified by:

z in interface Op

Returns:

the resulting ndarray

outputVariables

public SDVariable[] outputVariables(String baseName)

Description copied from class: DifferentialFunction

Return the output functions for this differential function.

Specified by:

outputVariables in class DifferentialFunction

Returns:

n

public long n()

Description copied from interface: Op

The number of elements to do a op over

Specified by:

n in interface Op

Returns:

the op

init

public void init(INDArray x,
                 INDArray y,
                 INDArray z,
                 long n)

Description copied from interface: Op

Initialize the operation based on the parameters

Specified by:

init in interface Op

Parameters:

x - the input

y - the pairwise transform ndarray

z - the resulting ndarray

n - the number of elements

setN

public void setN(long n)

Description copied from interface: Op

Change n

Specified by:

setN in interface Op

numProcessed

public long numProcessed()

Description copied from interface: Op

Number processed

Specified by:

numProcessed in interface Op

Returns:

the number of elements accumulated

toString

public String toString()

Overrides:

toString in class Object

toCustomOp

public CustomOp toCustomOp()

Description copied from interface: Op

Converts this op to be a CustomOp A CustomOp is a more flexible op meant for multiple inputs and outputs. The default implementation in BaseOp converts a simple op to a multi input/output operation by mapping the x and y on to inputs , the op opName and the z on to outputs.

Specified by:

toCustomOp in interface Op

Returns:

the equivalent CustomOp

exec

public void exec()

Description copied from interface: Op

Execute the op if its pass through (not needed most of the time)

Specified by:

exec in interface Op

exec

public void exec(int... dimensions)

Description copied from interface: Op

Exec along each dimension

Specified by:

exec in interface Op

Parameters:

dimensions - the dimensions to execute on

equals

public boolean equals(Object o)

Overrides:

equals in class DifferentialFunction

hashCode

public int hashCode()

Overrides:

hashCode in class DifferentialFunction