Class MaxGaussianOp

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

Inheritance

Object

MaxGaussianOp

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(new string[]{"max", "a", "b"}, typeof(Math), "Max", new Type[]{typeof(double), typeof(double)})]
[Quality(QualityBand.Stable)]
public static class MaxGaussianOp

Fields

ForceProper

Static flag to force a proper distribution

Declaration

public static bool ForceProper

Field Value

Type	Description
Boolean

Methods

AAverageConditional(Gaussian, Gaussian, Gaussian)

EP message to val1.

Declaration

public static Gaussian AAverageConditional(Gaussian max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the val1 argument.

Remarks

The outgoing message is a distribution matching the moments of val1 as the random arguments are varied. The formula is proj[p(val1) sum_(max,val2) p(max,val2) factor(max,val1,val2)]/p(val1).

Exceptions

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

AAverageConditional(Gaussian, Gaussian, Double)

EP message to val1.

Declaration

public static Gaussian AAverageConditional(Gaussian max, Gaussian a, double b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for val2.

Returns

Type	Description
Gaussian	The outgoing EP message to the val1 argument.

Remarks

The outgoing message is a distribution matching the moments of val1 as the random arguments are varied. The formula is proj[p(val1) sum_(max) p(max) factor(max,val1,val2)]/p(val1).

Exceptions

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

AAverageConditional(Double, Gaussian, Gaussian)

EP message to val1.

Declaration

public static Gaussian AAverageConditional(double max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the val1 argument.

Remarks

The outgoing message is a distribution matching the moments of val1 as the random arguments are varied. The formula is proj[p(val1) sum_(val2) p(val2) factor(max,val1,val2)]/p(val1).

Exceptions

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

AAverageConditional(Double, Gaussian, Double)

EP message to val1.

Declaration

public static Gaussian AAverageConditional(double max, Gaussian a, double b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for val2.

Returns

Type	Description
Gaussian	The outgoing EP message to the val1 argument.

Remarks

The outgoing message is the factor viewed as a function of val1 conditioned on the given values.

Exceptions

Type	Condition
ImproperMessageException	a is not a proper distribution.

AverageLogFactor(Double, Double, Double)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(double max, double a, double b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Double	a	Constant value for val1.
Double	b	Constant value for val2.

Returns

Type	Description
Double	Zero.

Remarks

The formula for the result is log(factor(max,val1,val2)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

BAverageConditional(Gaussian, Gaussian, Gaussian)

EP message to val2.

Declaration

public static Gaussian BAverageConditional(Gaussian max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the val2 argument.

Remarks

The outgoing message is a distribution matching the moments of val2 as the random arguments are varied. The formula is proj[p(val2) sum_(max,val1) p(max,val1) factor(max,val1,val2)]/p(val2).

Exceptions

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

BAverageConditional(Gaussian, Double, Gaussian)

EP message to val2.

Declaration

public static Gaussian BAverageConditional(Gaussian max, double a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Double	a	Constant value for val1.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the val2 argument.

Remarks

The outgoing message is a distribution matching the moments of val2 as the random arguments are varied. The formula is proj[p(val2) sum_(max) p(max) factor(max,val1,val2)]/p(val2).

Exceptions

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

BAverageConditional(Double, Gaussian, Gaussian)

EP message to val2.

Declaration

public static Gaussian BAverageConditional(double max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the val2 argument.

Remarks

The outgoing message is a distribution matching the moments of val2 as the random arguments are varied. The formula is proj[p(val2) sum_(val1) p(val1) factor(max,val1,val2)]/p(val2).

Exceptions

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

BAverageConditional(Double, Double, Gaussian)

EP message to val2.

Declaration

public static Gaussian BAverageConditional(double max, double a, Gaussian b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Double	a	Constant value for val1.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the val2 argument.

Remarks

The outgoing message is the factor viewed as a function of val2 conditioned on the given values.

Exceptions

Type	Condition
ImproperMessageException	b is not a proper distribution.

LogAverageFactor(Gaussian, Gaussian, Gaussian)

Evidence message for EP.

Declaration

public static double LogAverageFactor(Gaussian max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(max,val1,val2) p(max,val1,val2) factor(max,val1,val2)).

Exceptions

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

LogAverageFactor(Gaussian, Gaussian, Double)

Evidence message for EP.

Declaration

public static double LogAverageFactor(Gaussian max, Gaussian a, double b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for val2.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(max,val1) p(max,val1) factor(max,val1,val2)).

Exceptions

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

LogAverageFactor(Gaussian, Double, Gaussian)

Evidence message for EP.

Declaration

public static double LogAverageFactor(Gaussian max, double a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Double	a	Constant value for val1.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(max,val2) p(max,val2) factor(max,val1,val2)).

Exceptions

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

LogAverageFactor(Double, Gaussian, Gaussian)

Evidence message for EP.

Declaration

public static double LogAverageFactor(double max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(val1,val2) p(val1,val2) factor(max,val1,val2)).

Exceptions

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

LogAverageFactor(Double, Gaussian, Double)

Evidence message for EP.

Declaration

public static double LogAverageFactor(double max, Gaussian a, double b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for val2.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(val1) p(val1) factor(max,val1,val2)).

Exceptions

Type	Condition
ImproperMessageException	a is not a proper distribution.

LogAverageFactor(Double, Double, Gaussian)

Evidence message for EP.

Declaration

public static double LogAverageFactor(double max, double a, Gaussian b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Double	a	Constant value for val1.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(val2) p(val2) factor(max,val1,val2)).

Exceptions

Type	Condition
ImproperMessageException	b is not a proper distribution.

LogAverageFactor(Double, Double, Double)

Evidence message for EP.

Declaration

public static double LogAverageFactor(double max, double a, double b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Double	a	Constant value for val1.
Double	b	Constant value for val2.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(factor(max,val1,val2)).

LogEvidenceRatio(Gaussian, Gaussian, Gaussian)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(Gaussian max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. 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_(max,val1,val2) p(max,val1,val2) factor(max,val1,val2) / sum_max p(max) messageTo(max)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

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

LogEvidenceRatio(Gaussian, Gaussian, Double)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(Gaussian max, Gaussian a, double b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for val2.

Returns

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

Remarks

The formula for the result is log(sum_(max,val1) p(max,val1) factor(max,val1,val2) / sum_max p(max) messageTo(max)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

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

LogEvidenceRatio(Gaussian, Double, Gaussian)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(Gaussian max, double a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max. Must be a proper distribution. If uniform, the result will be uniform.
Double	a	Constant value for val1.
Gaussian	b	Incoming message from val2. 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_(max,val2) p(max,val2) factor(max,val1,val2) / sum_max p(max) messageTo(max)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

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

LogEvidenceRatio(Double, Gaussian, Gaussian)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(double max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. 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_(val1,val2) p(val1,val2) factor(max,val1,val2)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

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

LogEvidenceRatio(Double, Gaussian, Double)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(double max, Gaussian a, double b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for val2.

Returns

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

Remarks

The formula for the result is log(sum_(val1) p(val1) factor(max,val1,val2)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

Type	Condition
ImproperMessageException	a is not a proper distribution.

LogEvidenceRatio(Double, Double, Gaussian)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(double max, double a, Gaussian b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Double	a	Constant value for val1.
Gaussian	b	Incoming message from val2. 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_(val2) p(val2) factor(max,val1,val2)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

Type	Condition
ImproperMessageException	b is not a proper distribution.

LogEvidenceRatio(Double, Double, Double)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(double max, double a, double b)

Parameters

Type	Name	Description
Double	max	Constant value for max.
Double	a	Constant value for val1.
Double	b	Constant value for val2.

Returns

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

Remarks

The formula for the result is log(factor(max,val1,val2)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

MaxAverageConditional(Gaussian, Gaussian, Gaussian)

EP message to max.

Declaration

public static Gaussian MaxAverageConditional(Gaussian max, Gaussian a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the max argument.

Remarks

The outgoing message is a distribution matching the moments of max as the random arguments are varied. The formula is proj[p(max) sum_(val1,val2) p(val1,val2) factor(max,val1,val2)]/p(max).

Exceptions

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

MaxAverageConditional(Gaussian, Gaussian, Double)

EP message to max.

Declaration

public static Gaussian MaxAverageConditional(Gaussian max, Gaussian a, double b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max.
Gaussian	a	Incoming message from val1. Must be a proper distribution. If uniform, the result will be uniform.
Double	b	Constant value for val2.

Returns

Type	Description
Gaussian	The outgoing EP message to the max argument.

Remarks

The outgoing message is a distribution matching the moments of max as the random arguments are varied. The formula is proj[p(max) sum_(val1) p(val1) factor(max,val1,val2)]/p(max).

Exceptions

Type	Condition
ImproperMessageException	a is not a proper distribution.

MaxAverageConditional(Gaussian, Double, Gaussian)

EP message to max.

Declaration

public static Gaussian MaxAverageConditional(Gaussian max, double a, Gaussian b)

Parameters

Type	Name	Description
Gaussian	max	Incoming message from max.
Double	a	Constant value for val1.
Gaussian	b	Incoming message from val2. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Gaussian	The outgoing EP message to the max argument.

Remarks

The outgoing message is a distribution matching the moments of max as the random arguments are varied. The formula is proj[p(max) sum_(val2) p(val2) factor(max,val1,val2)]/p(max).

Exceptions

Type	Condition
ImproperMessageException	b is not a proper distribution.

MaxAverageConditionalInit()

Declaration

public static Gaussian MaxAverageConditionalInit()

Returns

Type	Description
Gaussian

Remarks