gov.sandia.cognition.statistics.bayesian
Class RejectionSampling.ScalarEstimator<ObservationType>

java.lang.Object
  extended by gov.sandia.cognition.statistics.bayesian.RejectionSampling.ScalarEstimator<ObservationType>
Type Parameters:
ObservationType - Type of observations to use.
Enclosing class:
RejectionSampling<ObservationType,ParameterType>

public static class RejectionSampling.ScalarEstimator<ObservationType>
extends Object

Routine for estimating the minimum scalar needed to envelop the conjunctive distribution.


Nested Class Summary
 class RejectionSampling.ScalarEstimator.MinimizerFunction
          Minimization function that measures the difference between the logarithm of the sampler function minus the logarithm of the conjunctive distribution.
 
Constructor Summary
RejectionSampling.ScalarEstimator(BayesianParameter<Double,? extends ProbabilityFunction<ObservationType>,? extends UnivariateProbabilityDensityFunction> conjunctive, Iterable<? extends ObservationType> data)
          Creates a new instance of ScalarEstimator
 
Method Summary
 double estimateScalarFactor(UnivariateProbabilityDensityFunction sampler)
          Estimates the minimum scalar needed for the sampler distribution to envelope the conjunctive distribution
 double logConjunctive(Double parameter)
          Computes the logarithm of the conjunctive likelihood for the given parameter
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

RejectionSampling.ScalarEstimator

public RejectionSampling.ScalarEstimator(BayesianParameter<Double,? extends ProbabilityFunction<ObservationType>,? extends UnivariateProbabilityDensityFunction> conjunctive,
                                         Iterable<? extends ObservationType> data)
Creates a new instance of ScalarEstimator

Parameters:
conjunctive - Defines the parameter that connects the conditional and prior distributions.
data - Data to consider
Method Detail

logConjunctive

public double logConjunctive(Double parameter)
Computes the logarithm of the conjunctive likelihood for the given parameter

Parameters:
parameter - Parameter to update.
Returns:
Logarithm of the conjunctive likelihood for the given parameter

estimateScalarFactor

public double estimateScalarFactor(UnivariateProbabilityDensityFunction sampler)
Estimates the minimum scalar needed for the sampler distribution to envelope the conjunctive distribution

Parameters:
sampler - Distribution from which we sample and envelop the conjunctive distribution.
Returns:
Minimum scalar needed for the sampler distribution to envelope the conjunctive distribution