Class ConstrainEqualRandomOp<TDomain>

Provides outgoing messages for EqualRandom<TDomain, TDistribution>(TDomain, TDistribution), given random arguments to the function.

Inheritance

Object

ConstrainEqualRandomOp<TDomain>

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(Constrain), "EqualRandom<,>", new Type[]{})]
[Quality(QualityBand.Mature)]
public static class ConstrainEqualRandomOp<TDomain>

Type Parameters

Name	Description
TDomain	The domain of the constrained variables.

Methods

AverageLogFactor<TDistribution>(TDistribution, TDistribution)

Evidence message for VMP.

Declaration

public static double AverageLogFactor<TDistribution>(TDistribution value, TDistribution dist)
    where TDistribution : CanGetAverageLog<TDistribution>

Parameters

Type	Name	Description
TDistribution	value	Incoming message from `value`.
TDistribution	dist	Incoming message from `dist`.

Returns

Type	Description
Double	Average of the factor's log-value across the given argument distributions.

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

The formula for the result is sum_(value,dist) p(value,dist) log(factor(value,dist)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

AverageLogFactor<TDistribution>(TDomain, TDistribution)

Evidence message for VMP.

Declaration

public static double AverageLogFactor<TDistribution>(TDomain value, TDistribution dist)
    where TDistribution : CanGetLogProb<TDomain>

Parameters

Type	Name	Description
TDomain	value	Incoming message from `value`.
TDistribution	dist	Incoming message from `dist`. Must be a proper distribution. If uniform, the result will be uniform.

Returns

Type	Description
Double	Average of the factor's log-value across the given argument distributions.

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

The formula for the result is sum_(value,dist) p(value,dist) log(factor(value,dist)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

Exceptions

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

LogAverageFactor<TDistribution>(TDistribution, TDistribution)

Evidence message for EP.

Declaration

public static double LogAverageFactor<TDistribution>(TDistribution value, TDistribution dist)
    where TDistribution : CanGetLogAverageOf<TDistribution>

Parameters

Type	Name	Description
TDistribution	value	Incoming message from `value`.
TDistribution	dist	Incoming message from `dist`.

Returns

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

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

The formula for the result is log(sum_(value,dist) p(value,dist) factor(value,dist)).

LogAverageFactor<TDistribution>(TDomain, TDistribution)

Evidence message for EP.

Declaration

public static double LogAverageFactor<TDistribution>(TDomain value, TDistribution dist)
    where TDistribution : CanGetLogProb<TDomain>

Parameters

Type	Name	Description
TDomain	value	Incoming message from `value`.
TDistribution	dist	Incoming message from `dist`. 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.

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

The formula for the result is log(sum_(value,dist) p(value,dist) factor(value,dist)).

Exceptions

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

LogEvidenceRatio<TDistribution>(TDistribution, TDistribution)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio<TDistribution>(TDistribution value, TDistribution dist)
    where TDistribution : CanGetLogAverageOf<TDistribution>

Parameters

Type	Name	Description
TDistribution	value	Incoming message from `value`.
TDistribution	dist	Incoming message from `dist`.

Returns

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

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

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

LogEvidenceRatio<TDistribution>(TDomain, TDistribution)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio<TDistribution>(TDomain value, TDistribution dist)
    where TDistribution : CanGetLogProb<TDomain>

Parameters

Type	Name	Description
TDomain	value	Incoming message from `value`.
TDistribution	dist	Incoming message from `dist`.

Returns

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

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

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

ValueAverageConditional<TDistribution>(TDistribution)

EP message to value.

Declaration

public static TDistribution ValueAverageConditional<TDistribution>(TDistribution dist)

Parameters

Type	Name	Description
TDistribution	dist	Incoming message from `dist`.

Returns

Type	Description
TDistribution	The outgoing EP message to the `value` argument.

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

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

ValueAverageLogarithm<TDistribution>(TDistribution)

VMP message to value.

Declaration

public static TDistribution ValueAverageLogarithm<TDistribution>(TDistribution dist)

Parameters

Type	Name	Description
TDistribution	dist	Incoming message from `dist`.

Returns

Type	Description
TDistribution	The outgoing VMP message to the `value` argument.

Type Parameters

Name	Description
TDistribution	The distribution over the constrained variables.

Remarks

The outgoing message is the exponential of the average log-factor value, where the average is over all arguments except value. The formula is exp(sum_(dist) p(dist) log(factor(value,dist))).