Class DiscreteFromDiscreteOp
Provides outgoing messages for Discrete(Int32, Matrix), given random arguments to the function.
Namespace: Microsoft.ML.Probabilistic.Factors
Assembly: Microsoft.ML.Probabilistic.dll
Syntax
public static class DiscreteFromDiscreteOp : object
Methods
LogAverageFactor(Discrete, Discrete, Matrix)
Evidence message for EP.
Declaration
public static double LogAverageFactor(Discrete sample, Discrete selector, Matrix probs)
Parameters
Type  Name  Description 

Discrete  sample  Incoming message from 
Discrete  selector  Incoming message from 
Matrix  probs  Constant value for 
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,selector) p(sample,selector) factor(sample,selector,probs))
.
LogEvidenceRatio(Discrete, Discrete, Matrix)
Evidence message for EP.
Declaration
public static double LogEvidenceRatio(Discrete sample, Discrete selector, Matrix probs)
Parameters
Type  Name  Description 

Discrete  sample  Incoming message from 
Discrete  selector  Incoming message from 
Matrix  probs  Constant value for 
Returns
Type  Description 

Double  Logarithm of the factor's contribution the EP model evidence. 
Remarks
The formula for the result is log(sum_(sample,selector) p(sample,selector) factor(sample,selector,probs) / sum_sample p(sample) messageTo(sample))
. Adding up these values across all factors and variables gives the logevidence estimate for EP.
SampleAverageConditional(Discrete, Matrix, Discrete)
EP message to sample
.
Declaration
public static Discrete SampleAverageConditional(Discrete selector, Matrix probs, Discrete result)
Parameters
Type  Name  Description 

Discrete  selector  Incoming message from 
Matrix  probs  Constant value for 
Discrete  result  Modified to contain the outgoing message. 
Returns
Type  Description 

Discrete 

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_(selector) p(selector) factor(sample,selector,probs)]/p(sample)
.
SelectorAverageConditional(Discrete, Matrix, Discrete)
EP message to selector
.
Declaration
public static Discrete SelectorAverageConditional(Discrete sample, Matrix probs, Discrete result)
Parameters
Type  Name  Description 

Discrete  sample  Incoming message from 
Matrix  probs  Constant value for 
Discrete  result  Modified to contain the outgoing message. 
Returns
Type  Description 

Discrete 

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