Class PlusGammaOp

Provides outgoing messages for Plus(Double, Double), given random arguments to the function.

Inheritance

Object

PlusGammaOp

Inherited Members

Object.Equals(Object)

Object.Equals(Object, Object)

Object.GetHashCode()

Object.GetType()

Object.MemberwiseClone()

Object.ReferenceEquals(Object, Object)

Object.ToString()

Namespace: Microsoft.ML.Probabilistic.Factors

Assembly: Microsoft.ML.Probabilistic.dll

Syntax

[FactorMethod(typeof(Factor), "Plus", new Type[]{typeof(double), typeof(double)}, Default = true)]
[Quality(QualityBand.Experimental)]
public static class PlusGammaOp

Methods

AAverageConditional(GammaPower, GammaPower, GammaPower, GammaPower, GammaPower)

Declaration

public static GammaPower AAverageConditional(GammaPower sum, GammaPower a, GammaPower b, GammaPower to_a, GammaPower to_b)

Parameters

Type	Name	Description
GammaPower	sum
GammaPower	a
GammaPower	b
GammaPower	to_a
GammaPower	to_b

Returns

Type	Description
GammaPower

AAverageConditional(GammaPower, Double)

EP message to A.

Declaration

public static GammaPower AAverageConditional(GammaPower sum, double b)

Parameters

Type	Name	Description
GammaPower	sum	Incoming message from `Sum`. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for `B`.

Returns

Type	Description
GammaPower	The outgoing EP message to the `A` argument.

Remarks

The outgoing message is a distribution matching the moments of A as the random arguments are varied. The formula is proj[p(A) sum_(Sum) p(Sum) factor(Sum,A,B)]/p(A).

Exceptions

Type	Condition
ImproperMessageException	`sum` is not a proper distribution.

AAverageConditional2(GammaPower, GammaPower, GammaPower, GammaPower)

Declaration

public static GammaPower AAverageConditional2(GammaPower sum, GammaPower a, GammaPower b, GammaPower result)

Parameters

Type	Name	Description
GammaPower	sum
GammaPower	a
GammaPower	b
GammaPower	result

Returns

Type	Description
GammaPower

BAverageConditional(GammaPower, GammaPower, GammaPower, GammaPower, GammaPower)

Declaration

public static GammaPower BAverageConditional(GammaPower sum, GammaPower a, GammaPower b, GammaPower to_a, GammaPower to_b)

Parameters

Type	Name	Description
GammaPower	sum
GammaPower	a
GammaPower	b
GammaPower	to_a
GammaPower	to_b

Returns

Type	Description
GammaPower

BAverageConditional(GammaPower, Double)

EP message to B.

Declaration

public static GammaPower BAverageConditional(GammaPower sum, double a)

Parameters

Type	Name	Description
GammaPower	sum	Incoming message from `Sum`. Must be a proper distribution. If uniform, the result will be uniform.
Double	a	Constant value for `A`.

Returns

Type	Description
GammaPower	The outgoing EP message to the `B` argument.

Remarks

The outgoing message is a distribution matching the moments of B as the random arguments are varied. The formula is proj[p(B) sum_(Sum) p(Sum) factor(Sum,A,B)]/p(B).

Exceptions

Type	Condition
ImproperMessageException	`sum` is not a proper distribution.

BAverageConditional2(GammaPower, GammaPower, GammaPower, GammaPower)

Declaration

public static GammaPower BAverageConditional2(GammaPower sum, GammaPower a, GammaPower b, GammaPower result)

Parameters

Type	Name	Description
GammaPower	sum
GammaPower	a
GammaPower	b
GammaPower	result

Returns

Type	Description
GammaPower

GammaPowerFromDerivLogZ(GammaPower, Double, Double)

Declaration

public static GammaPower GammaPowerFromDerivLogZ(GammaPower a, double dlogZ, double ddlogZ)

Parameters

Type	Name	Description
GammaPower	a
Double	dlogZ
Double	ddlogZ

Returns

Type	Description
GammaPower

GetDerivLogZ(GammaPower, GammaPower, Double, Double, Double, Double, out Double, out Double)

Gets the derivatives of the function that converts gamma parameters of toSum into the integral with sum.

Declaration

public static void GetDerivLogZ(GammaPower sum, GammaPower toSum, double ds, double dds, double dr, double ddr, out double dlogZ, out double ddlogZ)

Parameters

Type	Name	Description
GammaPower	sum
GammaPower	toSum
Double	ds
Double	dds
Double	dr
Double	ddr
Double	dlogZ
Double	ddlogZ

GetGammaDerivs(Double, Double, Double, Double, Double, Double, out Double, out Double, out Double, out Double)

Gets the derivatives of the function that converts moments into gamma parameters.

Declaration

public static void GetGammaDerivs(double mean, double dmean, double ddmean, double variance, double dvariance, double ddvariance, out double ds, out double dds, out double dr, out double ddr)

Parameters

Type	Name	Description
Double	mean
Double	dmean
Double	ddmean
Double	variance
Double	dvariance
Double	ddvariance
Double	ds
Double	dds
Double	dr
Double	ddr

GetGammaMomentDerivs(GammaPower, out Double, out Double, out Double, out Double, out Double, out Double)

Gets first and second derivatives of the moments with respect to the rate parameter of the distribution.

Declaration

public static void GetGammaMomentDerivs(GammaPower gammaPower, out double mean, out double dmean, out double ddmean, out double variance, out double dvariance, out double ddvariance)

Parameters

Type	Name	Description
GammaPower	gammaPower
Double	mean
Double	dmean
Double	ddmean
Double	variance
Double	dvariance
Double	ddvariance

GetInverseGammaDerivs(Double, Double, Double, Double, Double, Double, out Double, out Double, out Double, out Double)

Gets the derivatives of the function that converts moments into gamma parameters.

Declaration

public static void GetInverseGammaDerivs(double mean, double dmean, double ddmean, double variance, double dvariance, double ddvariance, out double ds, out double dds, out double dr, out double ddr)

Parameters

Type	Name	Description
Double	mean
Double	dmean
Double	ddmean
Double	variance
Double	dvariance
Double	ddvariance
Double	ds
Double	dds
Double	dr
Double	ddr

GetInverseGammaMomentDerivs(GammaPower, out Double, out Double, out Double, out Double, out Double, out Double)

Gets first and second derivatives of the moments with respect to the rate parameter of the distribution.

Declaration

public static void GetInverseGammaMomentDerivs(GammaPower gammaPower, out double mean, out double dmean, out double ddmean, out double variance, out double dvariance, out double ddvariance)

Parameters

Type	Name	Description
GammaPower	gammaPower
Double	mean
Double	dmean
Double	ddmean
Double	variance
Double	dvariance
Double	ddvariance

GetPosteriorMeanAndVariance(Gamma, Double, Double, out Double, out Double)

Declaration

public static void GetPosteriorMeanAndVariance(Gamma prior, double dlogZ, double ddlogZ, out double mean, out double variance)

Parameters

Type	Name	Description
Gamma	prior
Double	dlogZ
Double	ddlogZ
Double	mean
Double	variance

LogAverageFactor(GammaPower, GammaPower, GammaPower)

Declaration

public static double LogAverageFactor(GammaPower sum, GammaPower a, GammaPower b)

Parameters

Type	Name	Description
GammaPower	sum
GammaPower	a
GammaPower	b

Returns

Type	Description
Double

LogEvidenceRatio(GammaPower, GammaPower, GammaPower)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(GammaPower sum, GammaPower a, GammaPower b)

Parameters

Type	Name	Description
GammaPower	sum	Incoming message from `Sum`. Must be a proper distribution. If uniform, the result will be uniform.
GammaPower	a	Incoming message from `A`. Must be a proper distribution. If uniform, the result will be uniform.
GammaPower	b	Incoming message from `B`. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Double	Logarithm of the factor's contribution the EP model evidence.

Remarks

The formula for the result is log(sum_(Sum,A,B) p(Sum,A,B) factor(Sum,A,B) / sum_Sum p(Sum) messageTo(Sum)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

Type	Condition
ImproperMessageException	`sum` is not a proper distribution.
ImproperMessageException	`a` is not a proper distribution.
ImproperMessageException	`b` is not a proper distribution.

SumAverageConditional(GammaPower, GammaPower, GammaPower)

Declaration

public static GammaPower SumAverageConditional(GammaPower a, GammaPower b, GammaPower result)

Parameters

Type	Name	Description
GammaPower	a
GammaPower	b
GammaPower	result

Returns

Type	Description
GammaPower

SumAverageConditional(GammaPower, Double)

EP message to Sum.

Declaration

public static GammaPower SumAverageConditional(GammaPower a, double b)

Parameters

Type	Name	Description
GammaPower	a	Incoming message from `A`. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for `B`.

Returns

Type	Description
GammaPower	The outgoing EP message to the `Sum` argument.

Remarks

The outgoing message is a distribution matching the moments of Sum as the random arguments are varied. The formula is proj[p(Sum) sum_(A) p(A) factor(Sum,A,B)]/p(Sum).

Exceptions

Type	Condition
ImproperMessageException	`a` is not a proper distribution.

SumAverageConditional(Double, GammaPower)

EP message to Sum.

Declaration

public static GammaPower SumAverageConditional(double a, GammaPower b)

Parameters

Type	Name	Description
Double	a	Constant value for `A`.
GammaPower	b	Incoming message from `B`. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
GammaPower	The outgoing EP message to the `Sum` argument.

Remarks

The outgoing message is a distribution matching the moments of Sum as the random arguments are varied. The formula is proj[p(Sum) sum_(B) p(B) factor(Sum,A,B)]/p(Sum).

Exceptions

Type	Condition
ImproperMessageException	`b` is not a proper distribution.