Class StringFromArrayOp

Provides outgoing messages for StringFromArray(Char[]), given random arguments to the function.

Inheritance

Object

StringFromArrayOp

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), "StringFromArray", new Type[]{})]
[Quality(QualityBand.Experimental)]
public static class StringFromArrayOp

Methods

CharactersAverageConditional<TDiscreteCharList>(StringDistribution, IList<DiscreteChar>, TDiscreteCharList)

EP message to characters.

Declaration

public static TDiscreteCharList CharactersAverageConditional<TDiscreteCharList>(StringDistribution str, IList<DiscreteChar> characters, TDiscreteCharList result)
    where TDiscreteCharList : IList<DiscreteChar>

Parameters

Type	Name	Description
StringDistribution	str	Incoming message from `str`.
IList<DiscreteChar>	characters	Incoming message from `characters`.
TDiscreteCharList	result	Modified to contain the outgoing message.

Returns

Type	Description
TDiscreteCharList	`result`

Type Parameters

Name	Description
TDiscreteCharList	The type of an outgoing message to `chars`.

Remarks

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

LogEvidenceRatio(IList<DiscreteChar>, StringDistribution)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(IList<DiscreteChar> characters, StringDistribution str)

Parameters

Type	Name	Description
IList<DiscreteChar>	characters	Incoming message from `characters`.
StringDistribution	str	Incoming message from `str`.

Returns

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

Remarks

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

LogEvidenceRatio(IList<DiscreteChar>, String)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(IList<DiscreteChar> characters, string str)

Parameters

Type	Name	Description
IList<DiscreteChar>	characters	Incoming message from `characters`.
String	str	Constant value for `str`.

Returns

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

Remarks

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

StrAverageConditional(IList<DiscreteChar>)

EP message to str.

Declaration

public static StringDistribution StrAverageConditional(IList<DiscreteChar> characters)

Parameters

Type	Name	Description
IList<DiscreteChar>	characters	Incoming message from `characters`.

Returns

Type	Description
StringDistribution	The outgoing EP message to the `str` argument.

Remarks

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