org.nd4j.linalg.api.ops.impl.accum

Class Variance

java.lang.Object

org.nd4j.linalg.api.ops.BaseOp

org.nd4j.linalg.api.ops.BaseAccumulation

org.nd4j.linalg.api.ops.impl.accum.Variance

All Implemented Interfaces:

Accumulation, Op

Direct Known Subclasses:

StandardDeviation

public class Variance
extends BaseAccumulation

Variance with bias correction. Bias can either be divided by n or adjusted with: (currentResult - (pow(bias, 2.0) / n())) / (n() - 1.0);

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 double	bias
protected boolean	biasCorrected
protected double	mean

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

applyFinalTransform, finalResult, finalResultComplex, isComplex

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

extraArgs, extraArgz, n, numProcessed, passThrough, x, y, z

Constructor Summary

Constructors

Constructor and Description
Variance()
Variance(boolean biasCorrected)
Variance(INDArray x)
Variance(INDArray x, boolean biasCorrected)
Variance(INDArray x, INDArray y)
Variance(INDArray x, INDArray y, boolean biasCorrected)
Variance(INDArray x, INDArray y, INDArray z, long n)
Variance(INDArray x, INDArray y, INDArray z, long n, boolean biasCorrected)
Variance(INDArray x, INDArray y, long n)
Variance(INDArray x, INDArray y, long n, boolean biasCorrected)

Method Summary

Modifier and Type	Method and Description
double	calculateFinalResult(double accum, long n) Calculate the final result.
float	calculateFinalResult(float accum, long n)
double	combineSubResults(double first, double second) Combine sub-results, when the Accumulation operation is split and executed in parallel. Sometimes this is the same as the update operation (i.e., in sum operation), but often it is not (for example, in L2 norm, sqrt(1/n * sum_i x_i^2).
float	combineSubResults(float first, float second)
IComplexNumber	combineSubResults(IComplexNumber first, IComplexNumber second)
void	exec() Execute the op if its pass through (not needed most of the time)
void	exec(int... dimension) Exec along each dimension
double	getAndSetFinalResult(double accum) Get and set the final result. The final result is also stored in the Accumulation op itself (and can be retrieved by getFinalResult()) In some Accumulation operations, a final operation must be done on the accumulated result.
float	getAndSetFinalResult(float accum)
IComplexNumber	getAndSetFinalResult(IComplexNumber accum) Complex version of getAndSetFinalResult().
void	init(INDArray x, INDArray y, INDArray z, long n) Initialize the operation based on the parameters
boolean	isBiasCorrected()
boolean	isPassThrough() Returns whether the op should be executed or not (through the executioner)
String	name() The name of this operation
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.
double	op(double origin) Transform an individual element
float	op(float origin) Transform an individual element
Variance	opForDimension(int index, int... dimension) A copy of this operation for a particular dimension of the input
Op	opForDimension(int index, int dimension) A copy of this operation for a particular dimension of the input
int	opNum() An op number
void	setBiasCorrected(boolean biasCorrected)
double	update(double accum, double x) Do one accumulation update for a single-argument accumulation, given the current accumulation value and another value to be processed/accumulated
double	update(double accum, double x, double y) Do an accumulation update for a pair-wise (op(x,y)) accumulation, given the current accumulation value and a pair of values to be processed/accumulated
float	update(float accum, float x)
float	update(float accum, float x, float y)
IComplexNumber	update(IComplexNumber accum, double x) Complex update.
IComplexNumber	update(IComplexNumber accum, double x, double y) Complex update.
IComplexNumber	update(IComplexNumber accum, IComplexNumber x) Complex update.
IComplexNumber	update(IComplexNumber accum, IComplexNumber x, double y) Complex update.
IComplexNumber	update(IComplexNumber accum, IComplexNumber x, IComplexNumber y) Complex update.

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

applyFinalTransform, currentResult, getFinalResult, getFinalResultComplex, isComplexAccumulation, numProcessed, op, op, op, op, op, op, setApplyFinalTransform, setFinalResult, setFinalResultComplex, zeroComplex, zeroDouble, zeroFloat, zeroHalf

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

extraArgs, extraArgsBuff, extraArgsDataBuff, getOpType, isExecSpecial, n, setN, setX, setY, setZ, toString, x, y, z

Methods inherited from class java.lang.Object

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

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

extraArgs, extraArgsBuff, extraArgsDataBuff, isExecSpecial, n, setExtraArgs, setN, setX, setY, setZ, x, y, z

Field Detail

mean

protected double mean

bias

protected double bias

biasCorrected

protected boolean biasCorrected

Constructor Detail

Variance

public Variance()

Variance

public Variance(boolean biasCorrected)

Variance

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

Variance

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

Variance

public Variance(INDArray x)

Variance

public Variance(INDArray x,
                INDArray y)

Variance

public Variance(INDArray x,
                INDArray y,
                INDArray z,
                long n,
                boolean biasCorrected)

Variance

public Variance(INDArray x,
                INDArray y,
                long n,
                boolean biasCorrected)

Variance

public Variance(INDArray x,
                boolean biasCorrected)

Variance

public Variance(INDArray x,
                INDArray y,
                boolean biasCorrected)

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

Overrides:

noOp in class BaseAccumulation

Returns:

the no op version of the input

op

public double op(double origin)

Description copied from interface: Op

Transform an individual element

Specified by:

op in interface Op

Overrides:

op in class BaseAccumulation

Parameters:

origin - the origin element

Returns:

the new element

op

public float op(float origin)

Description copied from interface: Op

Transform an individual element

Specified by:

op in interface Op

Overrides:

op in class BaseAccumulation

Parameters:

origin - the origin element

Returns:

the new element

update

public double update(double accum,
                     double x)

Description copied from interface: Accumulation

Do one accumulation update for a single-argument accumulation, given the current accumulation value and another value to be processed/accumulated

Parameters:

accum - The current accumulation value

x - The next/new value to be processed/accumulated

Returns:

The updated accumulation value

update

public double update(double accum,
                     double x,
                     double y)

Description copied from interface: Accumulation

Do an accumulation update for a pair-wise (op(x,y)) accumulation, given the current accumulation value and a pair of values to be processed/accumulated

Parameters:

accum - The current accumulation value

x - The next/new x value to be processed/accumulated

y - The next/new y value to be processed/accumulated

Returns:

the updated accumulation value

update

public float update(float accum,
                    float x)

See Also:

Accumulation.update(double, double)

update

public float update(float accum,
                    float x,
                    float y)

See Also:

Accumulation.update(double, double, double)

update

public IComplexNumber update(IComplexNumber accum,
                             double x)

Description copied from interface: Accumulation

Complex update.

See Also:

Accumulation.update(double, double)

update

public IComplexNumber update(IComplexNumber accum,
                             double x,
                             double y)

Description copied from interface: Accumulation

Complex update.

See Also:

Accumulation.update(double, double, double)

update

public IComplexNumber update(IComplexNumber accum,
                             IComplexNumber x)

Description copied from interface: Accumulation

Complex update.

See Also:

Accumulation.update(double, double)

update

public IComplexNumber update(IComplexNumber accum,
                             IComplexNumber x,
                             IComplexNumber y)

Description copied from interface: Accumulation

Complex update.

See Also:

Accumulation.update(double, double, double)

update

public IComplexNumber update(IComplexNumber accum,
                             IComplexNumber x,
                             double y)

Description copied from interface: Accumulation

Complex update.

See Also:

Accumulation.update(double, double, double)

opNum

public int opNum()

Description copied from interface: Op

An op number

Returns:

name

public String name()

Description copied from interface: Op

The name of this operation

Returns:

the name of this operation

opForDimension

public Op opForDimension(int index,
                         int dimension)

Description copied from interface: Op

A copy of this operation for a particular dimension of the input

Parameters:

index - the index of the op to iterate over

dimension - the dimension to ge the input for

Returns:

the operation for that dimension

opForDimension

public Variance opForDimension(int index,
                               int... dimension)

Description copied from interface: Op

A copy of this operation for a particular dimension of the input

Parameters:

index - the index of the op to iterate over

dimension - the dimension to ge the input for

Returns:

the operation for that dimension

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

Overrides:

init in class BaseAccumulation

Parameters:

x - the input

y - the pairwise transform ndarray

z - the resulting ndarray

n - the number of elements

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

Overrides:

isPassThrough in class BaseOp

Returns:

true if the op is pass through false otherwise

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

Overrides:

exec in class BaseOp

exec

public void exec(int... dimension)

Description copied from interface: Op

Exec along each dimension

Specified by:

exec in interface Op

Overrides:

exec in class BaseOp

Parameters:

dimension - the dimensions to execute on

combineSubResults

public double combineSubResults(double first,
                                double second)

Description copied from interface: Accumulation

Combine sub-results, when the Accumulation operation is split and executed in parallel.
Sometimes this is the same as the update operation (i.e., in sum operation), but often it is not (for example, in L2 norm, sqrt(1/n * sum_i x_i^2).

Specified by:

combineSubResults in interface Accumulation

Overrides:

combineSubResults in class BaseAccumulation

Parameters:

first - The first sub-value to be combined

second - The second sub-value to be combined

Returns:

The combined result

combineSubResults

public float combineSubResults(float first,
                               float second)

Specified by:

combineSubResults in interface Accumulation

Overrides:

combineSubResults in class BaseAccumulation

See Also:

Accumulation.combineSubResults(double, double)

combineSubResults

public IComplexNumber combineSubResults(IComplexNumber first,
                                        IComplexNumber second)

Specified by:

combineSubResults in interface Accumulation

Overrides:

combineSubResults in class BaseAccumulation

See Also:

Accumulation.combineSubResults(double, double)

getAndSetFinalResult

public double getAndSetFinalResult(double accum)

Description copied from interface: Accumulation

Get and set the final result.
The final result is also stored in the Accumulation op itself (and can be retrieved by getFinalResult()) In some Accumulation operations, a final operation must be done on the accumulated result. For example, division by n in a mean operation: the accumulated result is the sum, whereas the final result is the sum/n.

Specified by:

getAndSetFinalResult in interface Accumulation

Overrides:

getAndSetFinalResult in class BaseAccumulation

Parameters:

accum - The accumulated result

Returns:

The final result

getAndSetFinalResult

public float getAndSetFinalResult(float accum)

Specified by:

getAndSetFinalResult in interface Accumulation

Overrides:

getAndSetFinalResult in class BaseAccumulation

See Also:

Accumulation.getAndSetFinalResult(double)

getAndSetFinalResult

public IComplexNumber getAndSetFinalResult(IComplexNumber accum)

Description copied from interface: Accumulation

Complex version of getAndSetFinalResult().

Specified by:

getAndSetFinalResult in interface Accumulation

Overrides:

getAndSetFinalResult in class BaseAccumulation

See Also:

Accumulation.getAndSetFinalResult(double)

calculateFinalResult

public double calculateFinalResult(double accum,
                                   long n)

Description copied from interface: Accumulation

Calculate the final result. unlike Accumulation.getAndSetFinalResult(double), the result is merely calculated and returned, not stored in the Accumulation op itself.

Specified by:

calculateFinalResult in interface Accumulation

Overrides:

calculateFinalResult in class BaseAccumulation

Parameters:

accum - The accumulated result

n - the number of elements accumulated

Returns:

calculated final result

calculateFinalResult

public float calculateFinalResult(float accum,
                                  long n)

Specified by:

calculateFinalResult in interface Accumulation

Overrides:

calculateFinalResult in class BaseAccumulation

See Also:

Accumulation.calculateFinalResult(double, long)

isBiasCorrected

public boolean isBiasCorrected()

setBiasCorrected

public void setBiasCorrected(boolean biasCorrected)