de.up.ling.irtg.script

Class EvaluateSamplingFromRules

java.lang.Object

de.up.ling.irtg.script.EvaluateSamplingFromRules

public class EvaluateSamplingFromRules
extends Object

This class can be used to evaluate adaptive importance sampling on a folder of tree automata.

Author:

christoph

Constructor Summary

Constructors

Constructor and Description
EvaluateSamplingFromRules()

Method Summary

Modifier and Type	Method and Description
static <Type> DoubleList	computeTargetFunction(TreeAutomaton<Type> ta, AdaptiveImportanceSampler.Configuration config) Computes the expected value for all the different evaluation rounds in a single repetition.
static void	main(String... args) Load all configuration parameters from a configuration file and evaluates the quality of predicting the uniform weighting on the automata.
static <Type> Pair<DoubleList,List<DoubleList>>	makeInside(TreeAutomaton<Type> ta, AdaptiveImportanceSampler.Configuration config, int repetitions, boolean printTracking) Computes the expected value average over all repetitions and then a list of expected values per round for each repetition.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

EvaluateSamplingFromRules

public EvaluateSamplingFromRules()

Method Detail

main

public static void main(String... args)
                 throws IOException

Load all configuration parameters from a configuration file and evaluates the quality of predicting the uniform weighting on the automata. This is the class used for the evaluation in our paper.

Parameters:

args -

Throws:

IOException

computeTargetFunction

public static <Type> DoubleList computeTargetFunction(TreeAutomaton<Type> ta,
                                                      AdaptiveImportanceSampler.Configuration config)

Computes the expected value for all the different evaluation rounds in a single repetition.

Type Parameters:

Type -

Parameters:

ta -

config -

Returns:

makeInside

public static <Type> Pair<DoubleList,List<DoubleList>> makeInside(TreeAutomaton<Type> ta,
                                                                  AdaptiveImportanceSampler.Configuration config,
                                                                  int repetitions,
                                                                  boolean printTracking)

Computes the expected value average over all repetitions and then a list of expected values per round for each repetition.

Type Parameters:

Type -

Parameters:

ta -

config -

repetitions -

printTracking -

Returns:

first element are the average expected value estimates for each round second element is a list for each repetition, which contains the estiamte for each round.