org.nd4j.autodiff.samediff

Class SDVariable

java.lang.Object

org.nd4j.autodiff.functions.DifferentialFunction

org.nd4j.autodiff.samediff.SDVariable

All Implemented Interfaces:

Serializable

public class SDVariable
extends DifferentialFunction
implements Serializable

A variable representing a component within a graph. SDVariable is used for symbolic declaration of equations.

Author:

Adam Gibson

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected WeightInitScheme	weightInitScheme

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

dimensions, extraArgs, inPlace, sameDiff, scalarValue

Method Summary

Modifier and Type	Method and Description
SDVariable	add(double sameDiffVariable)
SDVariable	add(SDVariable sameDiffVariable)
SDVariable	add(String varName, double sameDiffVariable)
SDVariable	add(String varName, SDVariable sameDiffVariable)
SDVariable	addi(double sameDiffVariable)
SDVariable	addi(SDVariable sameDiffVariable)
SDVariable	addi(String varName, double sameDiffVariable)
SDVariable	addi(String varName, SDVariable sameDiffVariable)
SDVariable	arg() Return the first argument
SDVariable[]	args() Return the arguments for a given function
SDVariable	div(double sameDiffVariable)
SDVariable	div(SDVariable sameDiffVariable)
SDVariable	div(String varName, double sameDiffVariable)
SDVariable	div(String varName, SDVariable sameDiffVariable)
SDVariable	divi(double sameDiffVariable)
SDVariable	divi(SDVariable sameDiffVariable)
SDVariable	divi(String varName, double sameDiffVariable)
SDVariable	divi(String varName, SDVariable sameDiffVariable)
List<SDVariable>	doDiff(List<SDVariable> f1) The actual implementation for automatic differentiation.
SDVariable	dup() Duplicate this function
boolean	equals(Object o)
INDArray	eval() Evaluate the result of this variable
INDArray	getArr() A getter for the allocated ndarray with this SDVariable.
SDVariable	getGradient() A getter for the variable gradient.
int[]	getShape() Returns the shape of this variable
SDVariable	gradient() Nicer looking alias for the gradient variable.
int	hashCode()
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	isPlaceHolder() Returns true if this variable is a place holder
SDVariable	mul(double sameDiffVariable)
SDVariable	mul(SDVariable sameDiffVariable)
SDVariable	mul(String varName, double sameDiffVariable)
SDVariable	mul(String varName, SDVariable sameDiffVariable)
SDVariable	muli(double sameDiffVariable)
SDVariable	muli(SDVariable sameDiffVariable)
SDVariable	muli(String varName, double sameDiffVariable)
SDVariable	muli(String varName, SDVariable sameDiffVariable)
String	onnxName() The opName of this function in onnx
String	opName() The name of the op
Op.Type	opType() The type of the op
SDVariable[]	outputVariables() Return the output variables for this differential function.
SDVariable[]	outputVariables(String baseName) Return the output functions for this differential function.
SDVariable	rdiv(double sameDiffVariable)
SDVariable	rdiv(SDVariable sameDiffVariable)
SDVariable	rdiv(String varName, double sameDiffVariable)
SDVariable	rdiv(String varName, SDVariable sameDiffVariable)
SDVariable	rdivi(double sameDiffVariable)
SDVariable	rdivi(SDVariable sameDiffVariable)
SDVariable	rdivi(String varName, double sameDiffVariable)
SDVariable	rdivi(String varName, SDVariable sameDiffVariable)
SDVariable	rsub(double sameDiffVariable)
SDVariable	rsub(SDVariable sameDiffVariable)
SDVariable	rsub(String varName, double sameDiffVariable)
SDVariable	rsub(String varName, SDVariable sameDiffVariable)
SDVariable	rsubi(double sameDiffVariable)
SDVariable	rsubi(SDVariable sameDiffVariable)
SDVariable	rsubi(String varName, double sameDiffVariable)
SDVariable	rsubi(String varName, SDVariable sameDiffVariable)
SDVariable	squaredDifference(SDVariable sameDiffVariable)
SDVariable	squaredDifference(String varName, SDVariable sameDiffVariable)
INDArray	storeAndAllocateNewArray() Allocate and return a new array based on the vertex id and weight initialization.
SDVariable	sub(double sameDiffVariable)
SDVariable	sub(SDVariable sameDiffVariable)
SDVariable	sub(String varName, double sameDiffVariable)
SDVariable	sub(String varName, SDVariable sameDiffVariable)
SDVariable	subi(double sameDiffVariable)
SDVariable	subi(SDVariable sameDiffVariable)
SDVariable	subi(String varName, double sameDiffVariable)
SDVariable	subi(String varName, SDVariable sameDiffVariable)
String	tensorflowName() The opName of this function tensorflow
String	toString()
SDVariable	truncatedDiv(SDVariable sameDiffVariable)
SDVariable	truncatedDiv(String varName, SDVariable sameDiffVariable)

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

asProperties, attributeAdaptersForFunction, calculateOutputShape, configFieldName, diff, f, getValue, hasPlaceHolderInputs, isConfigProperties, larg, mappingsForFunction, onnxNames, opNum, propertiesForFunction, rarg, resolvePropertiesFromSameDiffBeforeExecution, setInstanceId, setValueFor, tensorflowNames

Methods inherited from class java.lang.Object

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

Field Detail

weightInitScheme

protected WeightInitScheme weightInitScheme

Method Detail

isPlaceHolder

public boolean isPlaceHolder()

Returns true if this variable is a place holder

Returns:

opName

public String opName()

Description copied from class: DifferentialFunction

The name of the op

Overrides:

opName in class DifferentialFunction

Returns:

outputVariables

public SDVariable[] outputVariables()

Description copied from class: DifferentialFunction

Return the output variables for this differential function. Note that this op *may* dynamically generate variable outputs.

Overrides:

outputVariables in class DifferentialFunction

Returns:

arg

public SDVariable arg()

Description copied from class: DifferentialFunction

Return the first argument

Overrides:

arg in class DifferentialFunction

Returns:

args

public SDVariable[] args()

Description copied from class: DifferentialFunction

Return the arguments for a given function

Overrides:

args in class DifferentialFunction

Returns:

the arguments for a given function

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:

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

storeAndAllocateNewArray

public INDArray storeAndAllocateNewArray()

Allocate and return a new array based on the vertex id and weight initialization.

Returns:

the allocated array

getArr

public INDArray getArr()

A getter for the allocated ndarray with this SDVariable. This getter will lazy initialize an array if one is not found based on the associated shape and WeightInitScheme if neither are found, an ND4JIllegalStateException is thrown. If a DifferentialFunction is defined, note that its getArr() method is called instead.

Returns:

the INDArray associated with this variable.

gradient

public SDVariable gradient()

Nicer looking alias for the gradient variable. The gradient variable is meant to be an a variable representation of the gradient represented in the underlying DifferentialFunction

Returns:

getGradient

public SDVariable getGradient()

A getter for the variable gradient. Note here that a lazy initialization of the gradient variable will happen if the gradient isn't present at this variable's initialization but is set later.

Returns:

doDiff

public List<SDVariable> doDiff(List<SDVariable> f1)

Description copied from class: DifferentialFunction

The actual implementation for automatic differentiation.

Specified by:

doDiff in class DifferentialFunction

Returns:

getShape

public int[] getShape()

Returns the shape of this variable

Returns:

dup

public SDVariable dup()

Description copied from class: DifferentialFunction

Duplicate this function

Overrides:

dup in class DifferentialFunction

Returns:

rsub

public SDVariable rsub(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdiv

public SDVariable rdiv(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

add

public SDVariable add(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

sub

public SDVariable sub(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

squaredDifference

public SDVariable squaredDifference(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

div

public SDVariable div(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

mul

public SDVariable mul(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rsubi

public SDVariable rsubi(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdivi

public SDVariable rdivi(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

addi

public SDVariable addi(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

subi

public SDVariable subi(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

divi

public SDVariable divi(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

muli

public SDVariable muli(double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rsub

public SDVariable rsub(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdiv

public SDVariable rdiv(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

truncatedDiv

public SDVariable truncatedDiv(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

add

public SDVariable add(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

sub

public SDVariable sub(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

div

public SDVariable div(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

mul

public SDVariable mul(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rsubi

public SDVariable rsubi(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdivi

public SDVariable rdivi(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

addi

public SDVariable addi(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

subi

public SDVariable subi(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

divi

public SDVariable divi(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

muli

public SDVariable muli(SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rsub

public SDVariable rsub(String varName,
                       double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdiv

public SDVariable rdiv(String varName,
                       double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

truncatedDiv

public SDVariable truncatedDiv(String varName,
                               SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

add

public SDVariable add(String varName,
                      double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

sub

public SDVariable sub(String varName,
                      double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

div

public SDVariable div(String varName,
                      double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

mul

public SDVariable mul(String varName,
                      double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rsubi

public SDVariable rsubi(String varName,
                        double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdivi

public SDVariable rdivi(String varName,
                        double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

addi

public SDVariable addi(String varName,
                       double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

subi

public SDVariable subi(String varName,
                       double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

divi

public SDVariable divi(String varName,
                       double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

muli

public SDVariable muli(String varName,
                       double sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rsub

public SDVariable rsub(String varName,
                       SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdiv

public SDVariable rdiv(String varName,
                       SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

add

public SDVariable add(String varName,
                      SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

sub

public SDVariable sub(String varName,
                      SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

squaredDifference

public SDVariable squaredDifference(String varName,
                                    SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

squared difference between variables

div

public SDVariable div(String varName,
                      SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

mul

public SDVariable mul(String varName,
                      SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rsubi

public SDVariable rsubi(String varName,
                        SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

rdivi

public SDVariable rdivi(String varName,
                        SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

addi

public SDVariable addi(String varName,
                       SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

opType

public Op.Type opType()

Description copied from class: DifferentialFunction

The type of the op

Overrides:

opType in class DifferentialFunction

Returns:

subi

public SDVariable subi(String varName,
                       SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

divi

public SDVariable divi(String varName,
                       SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

muli

public SDVariable muli(String varName,
                       SDVariable sameDiffVariable)

Parameters:

sameDiffVariable -

Returns:

eval

public INDArray eval()

Evaluate the result of this variable

Returns:

toString

public String toString()

Overrides:

toString in class Object

equals

public boolean equals(Object o)

Overrides:

equals in class DifferentialFunction

hashCode

public int hashCode()

Overrides:

hashCode in class DifferentialFunction

onnxName

public String onnxName()

Description copied from class: DifferentialFunction

The opName of this function in onnx

Specified by:

onnxName in class DifferentialFunction

Returns:

tensorflowName

public String tensorflowName()

Description copied from class: DifferentialFunction

The opName of this function tensorflow

Specified by:

tensorflowName in class DifferentialFunction

Returns: