Class SparseBernoulliFromBetaOp

Provides outgoing messages for Sample(ISparseList<Double>), given random arguments to the function.

Inheritance

Object

SparseBernoulliFromBetaOp

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(SparseBernoulliList), "Sample", new Type[]{typeof(ISparseList<double>)})]
[Quality(QualityBand.Stable)]
public class SparseBernoulliFromBetaOp

Methods

AverageLogFactor(ISparseList<Boolean>, ISparseList<Double>)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(ISparseList<bool> sample, ISparseList<double> probTrue)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
ISparseList<Double>	probTrue	Constant value for probTrue.

Returns

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

Remarks

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

AverageLogFactor(ISparseList<Boolean>, SparseBetaList)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(ISparseList<bool> sample, SparseBetaList probTrue)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
SparseBetaList	probTrue	Incoming message from probTrue. Must be a proper distribution. If any element is uniform, the result will be uniform.

Returns

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

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	probTrue is not a proper distribution.

AverageLogFactor(SparseBernoulliList, ISparseList<Double>)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(SparseBernoulliList sample, ISparseList<double> probTrue)

Parameters

Type	Name	Description
SparseBernoulliList	sample	Incoming message from sample.
ISparseList<Double>	probTrue	Constant value for probTrue.

Returns

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

Remarks

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

AverageLogFactor(SparseBernoulliList, SparseBetaList)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(SparseBernoulliList sample, SparseBetaList probTrue)

Parameters

Type	Name	Description
SparseBernoulliList	sample	Incoming message from sample.
SparseBetaList	probTrue	Incoming message from probTrue. Must be a proper distribution. If any element is uniform, the result will be uniform.

Returns

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

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	probTrue is not a proper distribution.

LogAverageFactor(ISparseList<Boolean>, ISparseList<Double>)

Evidence message for EP.

Declaration

public static double LogAverageFactor(ISparseList<bool> sample, ISparseList<double> probTrue)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
ISparseList<Double>	probTrue	Constant value for probTrue.

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(sample,probTrue)).

LogAverageFactor(ISparseList<Boolean>, SparseBetaList)

Evidence message for EP.

Declaration

public static double LogAverageFactor(ISparseList<bool> sample, SparseBetaList probTrue)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
SparseBetaList	probTrue	Incoming message from probTrue.

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_(probTrue) p(probTrue) factor(sample,probTrue)).

LogAverageFactor(SparseBernoulliList, SparseBernoulliList)

Evidence message for EP.

Declaration

public static double LogAverageFactor(SparseBernoulliList sample, SparseBernoulliList to_sample)

Parameters

Type	Name	Description
SparseBernoulliList	sample	Incoming message from sample.
SparseBernoulliList	to_sample	Outgoing message to sample.

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_(sample) p(sample) factor(sample,probTrue)).

LogEvidenceRatio(ISparseList<Boolean>, ISparseList<Double>)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(ISparseList<bool> sample, ISparseList<double> probTrue)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
ISparseList<Double>	probTrue	Constant value for probTrue.

Returns

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

Remarks

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

LogEvidenceRatio(ISparseList<Boolean>, SparseBetaList)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(ISparseList<bool> sample, SparseBetaList probTrue)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
SparseBetaList	probTrue	Incoming message from probTrue.

Returns

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

Remarks

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

LogEvidenceRatio(SparseBernoulliList, ISparseList<Double>)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(SparseBernoulliList sample, ISparseList<double> probTrue)

Parameters

Type	Name	Description
SparseBernoulliList	sample	Incoming message from sample.
ISparseList<Double>	probTrue	Constant value for probTrue.

Returns

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

Remarks

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

LogEvidenceRatio(SparseBernoulliList, SparseBetaList)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(SparseBernoulliList sample, SparseBetaList probTrue)

Parameters

Type	Name	Description
SparseBernoulliList	sample	Incoming message from sample.
SparseBetaList	probTrue	Incoming message from probTrue.

Returns

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

Remarks

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

ProbTrueAverageConditional(ISparseList<Boolean>, SparseBetaList)

EP message to probTrue.

Declaration

public static SparseBetaList ProbTrueAverageConditional(ISparseList<bool> sample, SparseBetaList result)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
SparseBetaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBetaList	result

Remarks

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

ProbTrueAverageConditional(SparseBernoulliList, SparseBetaList, SparseBetaList)

EP message to probTrue.

Declaration

public static SparseBetaList ProbTrueAverageConditional(SparseBernoulliList sample, SparseBetaList probTrue, SparseBetaList result)

Parameters

Type	Name	Description
SparseBernoulliList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseBetaList	probTrue	Incoming message from probTrue.
SparseBetaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBetaList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.

ProbTrueAverageLogarithm(ISparseList<Boolean>, SparseBetaList)

VMP message to probTrue.

Declaration

public static SparseBetaList ProbTrueAverageLogarithm(ISparseList<bool> sample, SparseBetaList result)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
SparseBetaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBetaList	result

Remarks

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

ProbTrueAverageLogarithm(SparseBernoulliList, SparseBetaList)

VMP message to probTrue.

Declaration

public static SparseBetaList ProbTrueAverageLogarithm(SparseBernoulliList sample, SparseBetaList result)

Parameters

Type	Name	Description
SparseBernoulliList	sample	Incoming message from sample.
SparseBetaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBetaList	result

Remarks

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

ProbTrueConditional(ISparseList<Boolean>, SparseBetaList)

Gibbs message to probTrue.

Declaration

public static SparseBetaList ProbTrueConditional(ISparseList<bool> sample, SparseBetaList result)

Parameters

Type	Name	Description
ISparseList<Boolean>	sample	Constant value for sample.
SparseBetaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBetaList	result

Remarks

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

SampleAverageConditional(ISparseList<Double>, SparseBernoulliList)

EP message to sample.

Declaration

public static SparseBernoulliList SampleAverageConditional(ISparseList<double> probTrue, SparseBernoulliList result)

Parameters

Type	Name	Description
ISparseList<Double>	probTrue	Constant value for probTrue.
SparseBernoulliList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBernoulliList	result

Remarks

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

SampleAverageConditional(SparseBetaList, SparseBernoulliList)

EP message to sample.

Declaration

public static SparseBernoulliList SampleAverageConditional(SparseBetaList probTrue, SparseBernoulliList result)

Parameters

Type	Name	Description
SparseBetaList	probTrue	Incoming message from probTrue. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseBernoulliList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBernoulliList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	probTrue is not a proper distribution.

SampleAverageLogarithm(ISparseList<Double>, SparseBernoulliList)

VMP message to sample.

Declaration

public static SparseBernoulliList SampleAverageLogarithm(ISparseList<double> probTrue, SparseBernoulliList result)

Parameters

Type	Name	Description
ISparseList<Double>	probTrue	Constant value for probTrue.
SparseBernoulliList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBernoulliList	result

Remarks

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

SampleAverageLogarithm(SparseBetaList, SparseBernoulliList)

VMP message to sample.

Declaration

public static SparseBernoulliList SampleAverageLogarithm(SparseBetaList probTrue, SparseBernoulliList result)

Parameters

Type	Name	Description
SparseBetaList	probTrue	Incoming message from probTrue. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseBernoulliList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBernoulliList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	probTrue is not a proper distribution.

SampleConditional(ISparseList<Double>, SparseBernoulliList)

Gibbs message to sample.

Declaration

public static SparseBernoulliList SampleConditional(ISparseList<double> probTrue, SparseBernoulliList result)

Parameters

Type	Name	Description
ISparseList<Double>	probTrue	Constant value for probTrue.
SparseBernoulliList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseBernoulliList	result

Remarks

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