Class BinomialOp

Provides outgoing messages for Binomial(Int32, Double), given random arguments to the function.

Inheritance

Object

BinomialOp

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", "trialCount", "p"}, typeof(Rand), "Binomial", new Type[]{typeof(int), typeof(double)})]
[Quality(QualityBand.Preview)]
public static class BinomialOp

Remarks

The factor is f(sample,n,p) = choose(n,sample) p^sample (1-p)^(n-sample)

Methods

AverageLogFactor(Int32, Beta, Discrete)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(int sample, Beta p, Discrete trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Discrete	trialCount	Incoming message from n.

Returns

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

Remarks

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

AverageLogFactor(Int32, Beta, Int32)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(int sample, Beta p, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Int32	trialCount	Constant value for n.

Returns

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

Remarks

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

AverageLogFactor(Int32, Double, Discrete)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(int sample, double p, Discrete trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Discrete	trialCount	Incoming message from n.

Returns

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

Remarks

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

AverageLogFactor(Int32, Double, Int32)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(int sample, double p, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Int32	trialCount	Constant value for n.

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(binomial,n,p)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

LogAverageFactor(Discrete, Discrete)

Evidence message for EP.

Declaration

public static double LogAverageFactor(Discrete sample, Discrete to_sample)

Parameters

Type	Name	Description
Discrete	sample	Incoming message from binomial.
Discrete	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_(binomial) p(binomial) factor(binomial,n,p)).

LogAverageFactor(Int32, Beta, Discrete)

Evidence message for EP.

Declaration

public static double LogAverageFactor(int sample, Beta p, Discrete trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Discrete	trialCount	Incoming message from n.

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_(p,n) p(p,n) factor(binomial,n,p)).

LogAverageFactor(Int32, Beta, Int32)

Evidence message for EP.

Declaration

public static double LogAverageFactor(int sample, Beta p, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Int32	trialCount	Constant value for n.

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_(p) p(p) factor(binomial,n,p)).

LogAverageFactor(Int32, Double, Discrete)

Evidence message for EP.

Declaration

public static double LogAverageFactor(int sample, double p, Discrete trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Discrete	trialCount	Incoming message from n.

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_(n) p(n) factor(binomial,n,p)).

LogAverageFactor(Int32, Double, Int32)

Evidence message for EP.

Declaration

public static double LogAverageFactor(int sample, double p, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Int32	trialCount	Constant value for n.

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(binomial,n,p)).

LogEvidenceRatio(Discrete)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(Discrete sample)

Parameters

Type	Name	Description
Discrete	sample	Incoming message from binomial.

Returns

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

Remarks

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

LogEvidenceRatio(Int32, Beta, Discrete)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(int sample, Beta p, Discrete trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Discrete	trialCount	Incoming message from n.

Returns

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

Remarks

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

LogEvidenceRatio(Int32, Beta, Int32)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(int sample, Beta p, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Int32	trialCount	Constant value for n.

Returns

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

Remarks

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

LogEvidenceRatio(Int32, Double, Discrete)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(int sample, double p, Discrete trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Discrete	trialCount	Incoming message from n.

Returns

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

Remarks

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

LogEvidenceRatio(Int32, Double, Int32)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(int sample, double p, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Int32	trialCount	Constant value for n.

Returns

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

Remarks

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

PAverageConditional(Discrete, Int32)

EP message to p.

Declaration

public static Beta PAverageConditional(Discrete sample, int trialCount)

Parameters

Type	Name	Description
Discrete	sample	Incoming message from binomial.
Int32	trialCount	Constant value for n.

Returns

Type	Description
Beta	The outgoing EP message to the p argument.

Remarks

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

PAverageConditional(Int32, Discrete, Beta)

EP message to p.

Declaration

public static Beta PAverageConditional(int sample, Discrete trialCount, Beta p)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Discrete	trialCount	Incoming message from n.
Beta	p	Incoming message from p.

Returns

Type	Description
Beta	The outgoing EP message to the p argument.

Remarks

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

PAverageConditional(Int32, Int32)

EP message to p.

Declaration

public static Beta PAverageConditional(int sample, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Int32	trialCount	Constant value for n.

Returns

Type	Description
Beta	The outgoing EP message to the p argument.

Remarks

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

PAverageLogarithm(Discrete, Discrete)

VMP message to p.

Declaration

public static Beta PAverageLogarithm(Discrete sample, Discrete trialCount)

Parameters

Type	Name	Description
Discrete	sample	Incoming message from binomial.
Discrete	trialCount	Incoming message from n.

Returns

Type	Description
Beta	The outgoing VMP message to the p argument.

Remarks

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

PAverageLogarithm(Discrete, Int32)

VMP message to p.

Declaration

public static Beta PAverageLogarithm(Discrete sample, int trialCount)

Parameters

Type	Name	Description
Discrete	sample	Incoming message from binomial.
Int32	trialCount	Constant value for n.

Returns

Type	Description
Beta	The outgoing VMP message to the p argument.

Remarks

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

PAverageLogarithm(Int32, Discrete)

VMP message to p.

Declaration

public static Beta PAverageLogarithm(int sample, Discrete trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Discrete	trialCount	Incoming message from n.

Returns

Type	Description
Beta	The outgoing VMP message to the p argument.

Remarks

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

PAverageLogarithm(Int32, Int32)

VMP message to p.

Declaration

public static Beta PAverageLogarithm(int sample, int trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Int32	trialCount	Constant value for n.

Returns

Type	Description
Beta	The outgoing VMP message to the p argument.

Remarks

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

SampleAverageConditional(Beta, Discrete, Discrete)

EP message to binomial.

Declaration

public static Discrete SampleAverageConditional(Beta p, Discrete trialCount, Discrete result)

Parameters

Type	Name	Description
Beta	p	Incoming message from p.
Discrete	trialCount	Incoming message from n.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

SampleAverageConditional(Beta, Int32, Discrete)

EP message to binomial.

Declaration

public static Discrete SampleAverageConditional(Beta p, int trialCount, Discrete result)

Parameters

Type	Name	Description
Beta	p	Incoming message from p.
Int32	trialCount	Constant value for n.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

SampleAverageConditional(Double, Discrete, Discrete)

EP message to binomial.

Declaration

public static Discrete SampleAverageConditional(double p, Discrete trialCount, Discrete result)

Parameters

Type	Name	Description
Double	p	Constant value for p.
Discrete	trialCount	Incoming message from n.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

SampleAverageConditional(Double, Poisson)

EP message to binomial.

Declaration

public static Poisson SampleAverageConditional(double p, Poisson trialCount)

Parameters

Type	Name	Description
Double	p	Constant value for p.
Poisson	trialCount	Incoming message from n.

Returns

Type	Description
Poisson	The outgoing EP message to the binomial argument.

Remarks

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

SampleAverageConditional(Double, Int32, Discrete)

EP message to binomial.

Declaration

public static Discrete SampleAverageConditional(double p, int trialCount, Discrete result)

Parameters

Type	Name	Description
Double	p	Constant value for p.
Int32	trialCount	Constant value for n.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

SampleAverageConditionalInit(Discrete)

Declaration

public static Discrete SampleAverageConditionalInit(Discrete trialCount)

Parameters

Type	Name	Description
Discrete	trialCount	Incoming message from n.

Returns

Type	Description
Discrete

Remarks

SampleAverageConditionalInit(Int32)

Declaration

public static Discrete SampleAverageConditionalInit(int trialCount)

Parameters

Type	Name	Description
Int32	trialCount	Constant value for n.

Returns

Type	Description
Discrete

Remarks

TrialCountAverageConditional(Discrete, Beta, Discrete)

EP message to n.

Declaration

public static Discrete TrialCountAverageConditional(Discrete sample, Beta p, Discrete result)

Parameters

Type	Name	Description
Discrete	sample	Incoming message from binomial.
Beta	p	Incoming message from p.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

TrialCountAverageConditional(Discrete, Double, Discrete)

EP message to n.

Declaration

public static Discrete TrialCountAverageConditional(Discrete sample, double p, Discrete result)

Parameters

Type	Name	Description
Discrete	sample	Incoming message from binomial.
Double	p	Constant value for p.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

TrialCountAverageConditional(Poisson, Double)

EP message to n.

Declaration

public static Poisson TrialCountAverageConditional(Poisson sample, double p)

Parameters

Type	Name	Description
Poisson	sample	Incoming message from binomial.
Double	p	Constant value for p.

Returns

Type	Description
Poisson	The outgoing EP message to the n argument.

Remarks

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

TrialCountAverageConditional(Int32, Beta, Discrete)

EP message to n.

Declaration

public static Discrete TrialCountAverageConditional(int sample, Beta p, Discrete result)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

TrialCountAverageConditional(Int32, Double, Discrete)

EP message to n.

Declaration

public static Discrete TrialCountAverageConditional(int sample, double p, Discrete result)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

TrialCountAverageConditional(Int32, Double, Poisson)

EP message to n.

Declaration

public static Poisson TrialCountAverageConditional(int sample, double p, Poisson trialCount)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Poisson	trialCount	Incoming message from n.

Returns

Type	Description
Poisson	The outgoing EP message to the n argument.

Remarks

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

TrialCountAverageLogarithm(Int32, Beta, Discrete)

VMP message to n.

Declaration

public static Discrete TrialCountAverageLogarithm(int sample, Beta p, Discrete result)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Beta	p	Incoming message from p.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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

TrialCountAverageLogarithm(Int32, Double, Discrete)

VMP message to n.

Declaration

public static Discrete TrialCountAverageLogarithm(int sample, double p, Discrete result)

Parameters

Type	Name	Description
Int32	sample	Constant value for binomial.
Double	p	Constant value for p.
Discrete	result	Modified to contain the outgoing message.

Returns

Type	Description
Discrete	result

Remarks

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