Class SparseGaussianListOp

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

Inheritance

Object

SparseGaussianListOp

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[]{"sample", "mean", "precision"}, typeof(SparseGaussianList), "Sample", new Type[]{typeof(ISparseList<double>), typeof(ISparseList<double>)})]
[Quality(QualityBand.Stable)]
public static class SparseGaussianListOp

Methods

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

Evidence message for VMP.

Declaration

public static double AverageLogFactor(ISparseList<double> sample, ISparseList<double> mean, ISparseList<double> precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.

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,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

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

Evidence message for VMP.

Declaration

public static double AverageLogFactor(ISparseList<double> sample, ISparseList<double> mean, SparseGammaList precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. 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_(precs) p(precs) log(factor(sample,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

Exceptions

Type	Condition
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for VMP.

Declaration

public static double AverageLogFactor(ISparseList<double> sample, SparseGaussianList mean, ISparseList<double> precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.

Returns

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

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.

AverageLogFactor(ISparseList<Double>, SparseGaussianList, SparseGammaList)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(ISparseList<double> sample, SparseGaussianList mean, SparseGammaList precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. 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_(means,precs) p(means,precs) log(factor(sample,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for VMP.

Declaration

public static double AverageLogFactor(SparseGaussianList sample, ISparseList<double> mean, ISparseList<double> precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.

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,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.

AverageLogFactor(SparseGaussianList, ISparseList<Double>, SparseGammaList)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(SparseGaussianList sample, ISparseList<double> mean, SparseGammaList precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. 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,precs) p(sample,precs) log(factor(sample,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

AverageLogFactor(SparseGaussianList, SparseGaussianList, ISparseList<Double>)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(SparseGaussianList sample, SparseGaussianList mean, ISparseList<double> precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.

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,means) p(sample,means) log(factor(sample,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.

AverageLogFactor(SparseGaussianList, SparseGaussianList, SparseGammaList)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(SparseGaussianList sample, SparseGaussianList mean, SparseGammaList precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. 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,means,precs) p(sample,means,precs) log(factor(sample,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for EP.

Declaration

public static double LogAverageFactor(ISparseList<double> sample, ISparseList<double> mean, ISparseList<double> precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.

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,means,precs)).

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

Evidence message for EP.

Declaration

public static double LogAverageFactor(ISparseList<double> sample, ISparseList<double> mean, SparseGammaList precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is 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_(precs) p(precs) factor(sample,means,precs)).

Exceptions

Type	Condition
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for EP.

Declaration

public static double LogAverageFactor(ISparseList<double> sample, SparseGaussianList mean, ISparseList<double> precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.

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_(means) p(means) factor(sample,means,precs)).

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.

LogAverageFactor(ISparseList<Double>, SparseGaussianList, SparseGammaList, SparseGammaList)

Evidence message for EP.

Declaration

public static double LogAverageFactor(ISparseList<double> sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList to_precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.

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_(means,precs) p(means,precs) factor(sample,means,precs)).

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for EP.

Declaration

public static double LogAverageFactor(SparseGaussianList sample, ISparseList<double> mean, ISparseList<double> precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.

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,means,precs)).

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.

LogAverageFactor(SparseGaussianList, ISparseList<Double>, SparseGammaList, SparseGammaList)

Evidence message for EP.

Declaration

public static double LogAverageFactor(SparseGaussianList sample, ISparseList<double> mean, SparseGammaList precision, SparseGammaList to_precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.

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,precs) p(sample,precs) factor(sample,means,precs)).

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

LogAverageFactor(SparseGaussianList, SparseGaussianList, ISparseList<Double>)

Evidence message for EP.

Declaration

public static double LogAverageFactor(SparseGaussianList sample, SparseGaussianList mean, ISparseList<double> precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.

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,means) p(sample,means) factor(sample,means,precs)).

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.

LogAverageFactor(SparseGaussianList, SparseGaussianList, SparseGammaList, SparseGammaList)

Evidence message for EP.

Declaration

public static double LogAverageFactor(SparseGaussianList sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList to_precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.

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,means,precs) p(sample,means,precs) factor(sample,means,precs)).

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(ISparseList<double> sample, ISparseList<double> mean, ISparseList<double> precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.

Returns

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

Remarks

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

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

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(ISparseList<double> sample, ISparseList<double> mean, SparseGammaList precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is 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_(precs) p(precs) factor(sample,means,precs)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

Exceptions

Type	Condition
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(ISparseList<double> sample, SparseGaussianList mean, ISparseList<double> precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means.
ISparseList<Double>	precision	Constant value for precs.

Returns

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

Remarks

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

LogEvidenceRatio(ISparseList<Double>, SparseGaussianList, SparseGammaList, SparseGammaList)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(ISparseList<double> sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList to_precision)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.

Returns

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

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

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

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(SparseGaussianList sample, ISparseList<double> mean, ISparseList<double> precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample.
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.

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,means,precs) / sum_sample p(sample) messageTo(sample)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

LogEvidenceRatio(SparseGaussianList, ISparseList<Double>, SparseGammaList, SparseGaussianList, SparseGammaList)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(SparseGaussianList sample, ISparseList<double> mean, SparseGammaList precision, SparseGaussianList to_sample, SparseGammaList to_precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	to_sample	Outgoing message to sample.
SparseGammaList	to_precision	Previous outgoing message to precision.

Returns

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

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

LogEvidenceRatio(SparseGaussianList, SparseGaussianList, ISparseList<Double>)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(SparseGaussianList sample, SparseGaussianList mean, ISparseList<double> precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample.
SparseGaussianList	mean	Incoming message from means.
ISparseList<Double>	precision	Constant value for precs.

Returns

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

Remarks

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

LogEvidenceRatio(SparseGaussianList, SparseGaussianList, SparseGammaList, SparseGaussianList, SparseGammaList)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(SparseGaussianList sample, SparseGaussianList mean, SparseGammaList precision, SparseGaussianList to_sample, SparseGammaList to_precision)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	to_sample	Outgoing message to sample.
SparseGammaList	to_precision	Previous outgoing message to precision.

Returns

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

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

MeanAverageConditional(ISparseList<Double>, ISparseList<Double>, SparseGaussianList)

EP message to means.

Declaration

public static SparseGaussianList MeanAverageConditional(ISparseList<double> sample, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

MeanAverageConditional(ISparseList<Double>, SparseGaussianList, SparseGammaList, SparseGammaList, SparseGaussianList)

EP message to means.

Declaration

public static SparseGaussianList MeanAverageConditional(ISparseList<double> sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList to_precision, SparseGaussianList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

MeanAverageConditional(SparseGaussianList, ISparseList<Double>, SparseGaussianList)

EP message to means.

Declaration

public static SparseGaussianList MeanAverageConditional(SparseGaussianList sample, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.

MeanAverageConditional(SparseGaussianList, SparseGaussianList, SparseGammaList, SparseGammaList, SparseGaussianList)

EP message to means.

Declaration

public static SparseGaussianList MeanAverageConditional(SparseGaussianList sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList to_precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

MeanAverageLogarithm(ISparseList<Double>, ISparseList<Double>, SparseGaussianList)

VMP message to means.

Declaration

public static SparseGaussianList MeanAverageLogarithm(ISparseList<double> sample, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

MeanAverageLogarithm(ISparseList<Double>, SparseGammaList, SparseGaussianList)

VMP message to means.

Declaration

public static SparseGaussianList MeanAverageLogarithm(ISparseList<double> sample, SparseGammaList precision, SparseGaussianList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	precision is not a proper distribution.

MeanAverageLogarithm(SparseGaussianList, ISparseList<Double>, SparseGaussianList)

VMP message to means.

Declaration

public static SparseGaussianList MeanAverageLogarithm(SparseGaussianList sample, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.

MeanAverageLogarithm(SparseGaussianList, SparseGammaList, SparseGaussianList)

VMP message to means.

Declaration

public static SparseGaussianList MeanAverageLogarithm(SparseGaussianList sample, SparseGammaList precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

PrecisionAverageConditional(ISparseList<Double>, ISparseList<Double>, SparseGammaList)

EP message to precs.

Declaration

public static SparseGammaList PrecisionAverageConditional(ISparseList<double> sample, ISparseList<double> mean, SparseGammaList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

PrecisionAverageConditional(ISparseList<Double>, SparseGaussianList, SparseGammaList, SparseGammaList)

EP message to precs.

Declaration

public static SparseGammaList PrecisionAverageConditional(ISparseList<double> sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

PrecisionAverageConditional(SparseGaussianList, ISparseList<Double>, SparseGammaList, SparseGammaList)

EP message to precs.

Declaration

public static SparseGammaList PrecisionAverageConditional(SparseGaussianList sample, ISparseList<double> mean, SparseGammaList precision, SparseGammaList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

PrecisionAverageConditional(SparseGaussianList, SparseGaussianList, SparseGammaList, SparseGammaList)

EP message to precs.

Declaration

public static SparseGammaList PrecisionAverageConditional(SparseGaussianList sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

PrecisionAverageLogarithm(ISparseList<Double>, ISparseList<Double>, SparseGammaList)

VMP message to precs.

Declaration

public static SparseGammaList PrecisionAverageLogarithm(ISparseList<double> sample, ISparseList<double> mean, SparseGammaList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

PrecisionAverageLogarithm(ISparseList<Double>, SparseGaussianList, SparseGammaList)

VMP message to precs.

Declaration

public static SparseGammaList PrecisionAverageLogarithm(ISparseList<double> sample, SparseGaussianList mean, SparseGammaList result)

Parameters

Type	Name	Description
ISparseList<Double>	sample	Constant value for sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.

PrecisionAverageLogarithm(SparseGaussianList, ISparseList<Double>, SparseGammaList)

VMP message to precs.

Declaration

public static SparseGammaList PrecisionAverageLogarithm(SparseGaussianList sample, ISparseList<double> mean, SparseGammaList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.

PrecisionAverageLogarithm(SparseGaussianList, SparseGaussianList, SparseGammaList)

VMP message to precs.

Declaration

public static SparseGammaList PrecisionAverageLogarithm(SparseGaussianList sample, SparseGaussianList mean, SparseGammaList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGammaList	result

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	sample is not a proper distribution.
ImproperMessageException	mean is not a proper distribution.

SampleAverageConditional(ISparseList<Double>, ISparseList<Double>, SparseGaussianList)

EP message to sample.

Declaration

public static SparseGaussianList SampleAverageConditional(ISparseList<double> mean, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

SampleAverageConditional(SparseGaussianList, ISparseList<Double>, SparseGammaList, SparseGammaList, SparseGaussianList)

EP message to sample.

Declaration

public static SparseGaussianList SampleAverageConditional(SparseGaussianList sample, ISparseList<double> mean, SparseGammaList precision, SparseGammaList to_precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample.
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	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_(precs) p(precs) factor(sample,means,precs)]/p(sample).

Exceptions

Type	Condition
ImproperMessageException	precision is not a proper distribution.

SampleAverageConditional(SparseGaussianList, ISparseList<Double>, SparseGaussianList)

EP message to sample.

Declaration

public static SparseGaussianList SampleAverageConditional(SparseGaussianList mean, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	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_(means) p(means) factor(sample,means,precs)]/p(sample).

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.

SampleAverageConditional(SparseGaussianList, SparseGaussianList, SparseGammaList, SparseGammaList, SparseGaussianList)

EP message to sample.

Declaration

public static SparseGaussianList SampleAverageConditional(SparseGaussianList sample, SparseGaussianList mean, SparseGammaList precision, SparseGammaList to_precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	sample	Incoming message from sample.
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	to_precision	Previous outgoing message to precision.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	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_(means,precs) p(means,precs) factor(sample,means,precs)]/p(sample).

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

SampleAverageConditionalInit(ISparseList<Double>)

Declaration

public static SparseGaussianList SampleAverageConditionalInit(ISparseList<double> mean)

Parameters

Type	Name	Description
ISparseList<Double>	mean	Constant value for means.

Returns

Type	Description
SparseGaussianList

Remarks

SampleAverageConditionalInit(SparseGaussianList)

Declaration

public static SparseGaussianList SampleAverageConditionalInit(SparseGaussianList mean)

Parameters

Type	Name	Description
SparseGaussianList	mean	Incoming message from means.

Returns

Type	Description
SparseGaussianList

Remarks

SampleAverageLogarithm(ISparseList<Double>, ISparseList<Double>, SparseGaussianList)

VMP message to sample.

Declaration

public static SparseGaussianList SampleAverageLogarithm(ISparseList<double> mean, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
ISparseList<Double>	mean	Constant value for means.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	result

Remarks

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

SampleAverageLogarithm(ISparseList<Double>, SparseGammaList, SparseGaussianList)

VMP message to sample.

Declaration

public static SparseGaussianList SampleAverageLogarithm(ISparseList<double> mean, SparseGammaList precision, SparseGaussianList result)

Parameters

Type	Name	Description
ISparseList<Double>	mean	Constant value for means.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	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_(precs) p(precs) log(factor(sample,means,precs))).

Exceptions

Type	Condition
ImproperMessageException	precision is not a proper distribution.

SampleAverageLogarithm(SparseGaussianList, ISparseList<Double>, SparseGaussianList)

VMP message to sample.

Declaration

public static SparseGaussianList SampleAverageLogarithm(SparseGaussianList mean, ISparseList<double> precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
ISparseList<Double>	precision	Constant value for precs.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	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_(means) p(means) log(factor(sample,means,precs))).

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.

SampleAverageLogarithm(SparseGaussianList, SparseGammaList, SparseGaussianList)

VMP message to sample.

Declaration

public static SparseGaussianList SampleAverageLogarithm(SparseGaussianList mean, SparseGammaList precision, SparseGaussianList result)

Parameters

Type	Name	Description
SparseGaussianList	mean	Incoming message from means. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGammaList	precision	Incoming message from precs. Must be a proper distribution. If any element is uniform, the result will be uniform.
SparseGaussianList	result	Modified to contain the outgoing message.

Returns

Type	Description
SparseGaussianList	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_(means,precs) p(means,precs) log(factor(sample,means,precs))).

Exceptions

Type	Condition
ImproperMessageException	mean is not a proper distribution.
ImproperMessageException	precision is not a proper distribution.

SampleAverageLogarithmInit(ISparseList<Double>)

Declaration

public static SparseGaussianList SampleAverageLogarithmInit(ISparseList<double> mean)

Parameters

Type	Name	Description
ISparseList<Double>	mean	Constant value for means.

Returns

Type	Description
SparseGaussianList

Remarks

SampleAverageLogarithmInit(SparseGaussianList)

Declaration

public static SparseGaussianList SampleAverageLogarithmInit(SparseGaussianList mean)

Parameters

Type	Name	Description
SparseGaussianList	mean	Incoming message from means.

Returns

Type	Description
SparseGaussianList

Remarks