Class UsesEqualDefOp

Provides outgoing messages for UsesEqualDef<T>(T, Int32, out T), given random arguments to the function.

Inheritance

Object

UsesEqualDefOp

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(Clone), "UsesEqualDef<>", new Type[]{})]
[Quality(QualityBand.Mature)]
public static class UsesEqualDefOp

Methods

DefAverageConditional<T>(IReadOnlyList<T>, T)

Declaration

public static T DefAverageConditional<T>([SkipIfAllUniform] IReadOnlyList<T> Uses, T result)
    where T : SettableToProduct<T>, SettableTo<T>, SettableToUniform

Parameters

Type	Name	Description
IReadOnlyList<T>	Uses
T	result

Returns

Type	Description
T

Type Parameters

Name	Description
T	The type of the messages.

LogEvidenceRatio<T>(IList<T>, T, IList<T>)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio<T>([SkipIfAllUniform] IList<T> Uses, T Def, IList<T> to_Uses)
    where T : CanGetLogAverageOf<T>, SettableToProduct<T>, SettableTo<T>, ICloneable, SettableToUniform

Parameters

Type	Name	Description
IList<T>	Uses	Incoming message from Uses. Must be a proper distribution. If all elements are uniform, the result will be uniform.
T	Def	Incoming message from Def.
IList<T>	to_Uses	Outgoing message to Uses.

Returns

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

Type Parameters

Name	Description
T	The type of the messages.

Remarks

The formula for the result is log(sum_(Uses,Def) p(Uses,Def) factor(Uses,Def,count,Marginal) / sum_Uses p(Uses) messageTo(Uses)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

Type	Condition
ImproperMessageException	Uses is not a proper distribution.

LogEvidenceRatio1<T>(IList<T>, T)

Evidence message for EP

Declaration

public static double LogEvidenceRatio1<T>([SkipIfAllUniform] IList<T> Uses, T Def)
    where T : CanGetLogAverageOf<T>, SettableToProduct<T>, SettableTo<T>, ICloneable, SettableToUniform

Parameters

Type	Name	Description
IList<T>	Uses	Incoming message from 'Uses'. Must be a proper distribution. If all elements are uniform, the result will be uniform.
T	Def	Incoming message from 'Def'.

Returns

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

Type Parameters

Name	Description
T

Remarks

The formula for the result is log(sum_(Uses,Def) p(Uses,Def) factor(Uses,Def,Marginal) / sum_Uses p(Uses) messageTo(Uses)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

Type	Condition
ImproperMessageException	Uses is not a proper distribution

MarginalAverageConditional<T>(T[], T, T)

EP message to Marginal.

Declaration

[SkipIfAllUniform]
public static T MarginalAverageConditional<T>(T[] Uses, T Def, T result)
    where T : SettableToProduct<T>, SettableTo<T>

Parameters

Type	Name	Description
T[]	Uses	Incoming message from Uses.
T	Def	Incoming message from Def.
T	result	Modified to contain the outgoing message.

Returns

Type	Description
T	result

Type Parameters

Name	Description
T	The type of the messages.

Remarks

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

MarginalAverageConditional<T>(IReadOnlyList<T>, T, T)

Declaration

public static T MarginalAverageConditional<T>(IReadOnlyList<T> Uses, T Def, T result)
    where T : SettableToProduct<T>, SettableTo<T>

Parameters

Type	Name	Description
IReadOnlyList<T>	Uses
T	Def
T	result

Returns

Type	Description
T

Type Parameters

Name	Description
T	The type of the messages.

UsesAverageConditional<T>(T[], T, Int32, T)

EP message to Uses.

Declaration

public static T UsesAverageConditional<T>(T[] Uses, [SkipIfAllUniform] T Def, int resultIndex, T result)
    where T : SettableToProduct<T>, SettableTo<T>

Parameters

Type	Name	Description
T[]	Uses	Incoming message from Uses.
T	Def	Incoming message from Def. Must be a proper distribution. If uniform, the result will be uniform.
Int32	resultIndex	Index of the Uses for which a message is desired.
T	result	Modified to contain the outgoing message.

Returns

Type	Description
T	result

Type Parameters

Name	Description
T	The type of the messages.

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	Def is not a proper distribution.

UsesAverageConditional<T>(IReadOnlyList<T>, T, Int32, T)

Declaration

public static T UsesAverageConditional<T>(IReadOnlyList<T> Uses, [SkipIfAllUniform] T Def, int resultIndex, T result)
    where T : SettableToProduct<T>, SettableTo<T>

Parameters

Type	Name	Description
IReadOnlyList<T>	Uses
T	Def
Int32	resultIndex
T	result

Returns

Type	Description
T

Type Parameters

Name	Description
T	The type of the messages.