joshua/hybrid-ads
1
Fork 0
mirror of https://github.com/Jaxan/hybrid-ads.git synced 2025-04-27 23:17:44 +02:00
Code Issues Releases Activity

Incorporates some changes by Wouter.

Browse source 
Makes main more dynamic and easy to adjust.
This commit is contained in:
Joshua Moerman 2015-04-10 16:08:11 +02:00
parent f13d8a0dd6
commit ba868d39c5
1 changed files with 91 additions and 52 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

137
java/Main.java
View file

@ -1,27 +1,22 @@
package yannakakis; package yannakakis;
import com.google.common.base.Supplier;
import com.google.common.collect.Lists; import com.google.common.collect.Lists;
import de.learnlib.acex.analyzers.AcexAnalyzers; import de.learnlib.acex.analyzers.AcexAnalyzers;
import de.learnlib.algorithms.dhc.mealy.MealyDHC;
import de.learnlib.algorithms.kv.mealy.KearnsVaziraniMealy; import de.learnlib.algorithms.kv.mealy.KearnsVaziraniMealy;
import de.learnlib.algorithms.lstargeneric.ce.ObservationTableCEXHandlers; import de.learnlib.algorithms.lstargeneric.ce.ObservationTableCEXHandlers;
import de.learnlib.algorithms.lstargeneric.closing.ClosingStrategies; import de.learnlib.algorithms.lstargeneric.closing.ClosingStrategies;
import de.learnlib.algorithms.lstargeneric.mealy.ExtensibleLStarMealy; import de.learnlib.algorithms.lstargeneric.mealy.ExtensibleLStarMealy;
import de.learnlib.algorithms.ttt.mealy.TTTLearnerMealy; import de.learnlib.algorithms.ttt.mealy.TTTLearnerMealy;
import de.learnlib.algorithms.ttt.mealy.TTTLearnerMealyBuilder;
import de.learnlib.api.EquivalenceOracle; import de.learnlib.api.EquivalenceOracle;
import de.learnlib.api.LearningAlgorithm; import de.learnlib.api.LearningAlgorithm;
import de.learnlib.api.MembershipOracle; import de.learnlib.api.MembershipOracle;
import de.learnlib.api.Query; import de.learnlib.api.Query;
import de.learnlib.counterexamples.LocalSuffixFinders; import de.learnlib.counterexamples.LocalSuffixFinders;
import de.learnlib.eqtests.basic.EQOracleChain;
import de.learnlib.eqtests.basic.SimulatorEQOracle;
import de.learnlib.eqtests.basic.WMethodEQOracle; import de.learnlib.eqtests.basic.WMethodEQOracle;
import de.learnlib.eqtests.basic.WpMethodEQOracle; import de.learnlib.eqtests.basic.WpMethodEQOracle;
import de.learnlib.oracles.AbstractSingleQueryOracle;
import de.learnlib.oracles.DefaultQuery; import de.learnlib.oracles.DefaultQuery;
import de.learnlib.oracles.SimulatorOracle; import de.learnlib.oracles.SimulatorOracle;
import de.learnlib.parallelism.DynamicParallelOracle;
import de.learnlib.parallelism.ParallelOracle; import de.learnlib.parallelism.ParallelOracle;
import de.learnlib.parallelism.ParallelOracleBuilders; import de.learnlib.parallelism.ParallelOracleBuilders;
import net.automatalib.automata.transout.MealyMachine; import net.automatalib.automata.transout.MealyMachine;
@ -29,73 +24,115 @@ import net.automatalib.automata.transout.impl.compact.CompactMealy;
import net.automatalib.util.graphs.dot.GraphDOT; import net.automatalib.util.graphs.dot.GraphDOT;
import net.automatalib.words.Alphabet; import net.automatalib.words.Alphabet;
import net.automatalib.words.Word; import net.automatalib.words.Word;
import net.automatalib.words.WordBuilder;
import net.automatalib.words.impl.Alphabets; import net.automatalib.words.impl.Alphabets;
import javax.annotation.Nullable;
import java.io.IOException; import java.io.IOException;
import java.io.PrintStream; import java.io.PrintStream;
import java.io.PrintWriter; import java.io.PrintWriter;
import java.nio.file.Paths; import java.nio.file.Paths;
import java.util.*; import java.util.Arrays;
import java.util.function.Consumer; import java.util.Calendar;
import java.util.function.Supplier; import java.util.Collection;
import java.util.List;
import static de.learnlib.cache.mealy.MealyCaches.createCache;
/** /**
* Test for the Lee and Yannakakis implementation. * Test for the Lee and Yannakakis implementation.
*/ */
public class Main { public class Main {
public static enum Method { LeeYannakakis, WMethod, WpMethod }
public static enum Learner { LStar, RivestSchapire, TTT, KearnsVazirani }
public static void main(String[] args) throws IOException { public static void main(String[] args) throws IOException {
// We read the mealy machine we want to simulate // Some settings
boolean learnRealESM = false;
boolean saveAllHypotheses = false;
int threadsForMembershipQueries = 4;
Method testMethod = Method.LeeYannakakis;
int maxDepthForWMethod = 2;
Learner learnAlgorithm = Learner.LStar;
System.out.println("Setting up SUL");
Supplier<MembershipOracle.MealyMembershipOracle<String, String>> supplier = null;
Alphabet<String> alphabet = null;
List<Alphabet<String>> alphabets = null;
if(learnRealESM){
// Real ESM
// I did not have this class, so I commented this out.
// long seed = 78465782;
// Random random = new Random(seed);
// supplier = new ESMOracleSupplier(random);
//
// alphabet = esmSupplier.getInputAlphabet();
// Alphabet<String> realSubAlphabet = esmSupplier.getSubAlphabet();
// alphabets = Arrays.asList(realSubAlphabet, realAlphabet);
} else {
// A simulated ESM (by means of a mealy machine)
String filename = "esm-manual-controller.dot"; String filename = "esm-manual-controller.dot";
System.out.println("Reading dot file: " + filename); System.out.println(" Reading dot file: " + filename);
GraphvizParser p = new GraphvizParser(Paths.get(filename)); GraphvizParser p = new GraphvizParser(Paths.get(filename));
CompactMealy<String, String> fm = p.createMachine(); CompactMealy<String, String> fm = p.createMachine();
Alphabet<String> alphabet = fm.getInputAlphabet(); alphabet = fm.getInputAlphabet();
Alphabet<String> subAlphabet = Alphabets.fromArray("21.1", "21.0", "22", "53.4", "52.5"); Alphabet<String> simSubAlphabet = Alphabets.fromArray("21.1", "21.0", "22", "53.4", "52.5");
List<Alphabet<String>> alphabets = Arrays.asList(subAlphabet, alphabet); alphabets = Arrays.asList(simSubAlphabet, alphabet);
System.out.println("created machine with " + fm.size() + " states and " + alphabet.size() + " inputs\n"); System.out.println(" created machine with " + fm.size() + " states and " + alphabet.size() + " inputs");
supplier = () -> {
return new SimulatorOracle.MealySimulatorOracle<>(fm);
};
}
assert supplier != null;
assert alphabet != null;
assert alphabets != null;
System.out.println("Setting up membership oracle");
MembershipOracle<String, Word<String>> mParallelOracle = ParallelOracleBuilders
.newDynamicParallelOracle(supplier)
.withBatchSize(5)
.withPoolSize(threadsForMembershipQueries)
.withPoolPolicy(ParallelOracle.PoolPolicy.FIXED)
.create();
// We will have the simulating membership oracle. Since the equivalence oracles use a membership oracle to // We will have the simulating membership oracle. Since the equivalence oracles use a membership oracle to
// test equality, we make membership oracle which maintain a count. E.g. This way we can count the number of // test equality, we make membership oracle which maintain a count. E.g. This way we can count the number of
// queries needed to find a counterexample. // queries needed to find a counterexample.
System.out.println("Setting up membership oracles");
SimulatorOracle.MealySimulatorOracle<String, String> mOracleMealy = new SimulatorOracle.MealySimulatorOracle<>(fm);
DynamicParallelOracle<String, Word<String>> mParallelOracle = ParallelOracleBuilders.newDynamicParallelOracle(() -> {
return new SimulatorOracle.MealySimulatorOracle<>(fm);
}).withBatchSize(5)
.withPoolSize(4)
.withPoolPolicy(ParallelOracle.PoolPolicy.FIXED)
.create();
CountingMembershipOracle<String, Word<String>> mOracleForLearning = new CountingMembershipOracle<>(mParallelOracle, "learning"); CountingMembershipOracle<String, Word<String>> mOracleForLearning = new CountingMembershipOracle<>(mParallelOracle, "learning");
CountingMembershipOracle<String, Word<String>> mOracleForTesting = new CountingMembershipOracle<>(mParallelOracle, "testing"); CountingMembershipOracle<String, Word<String>> mOracleForTesting = new CountingMembershipOracle<>(mParallelOracle, "testing");
// We can test multiple equivalence oracles. They all treat the SUL as a black box. System.out.println("Setting up equivalence oracle");
System.out.println("Setting up equivalence oracles"); EquivalenceOracle<MealyMachine<?, String, ?, String>, String, Word<String>> eqOracle = null;
WpMethodEQOracle.MealyWpMethodEQOracle<String, String> eqOracleWp = new WpMethodEQOracle.MealyWpMethodEQOracle<>(3, mOracleForTesting); switch (testMethod){
WMethodEQOracle.MealyWMethodEQOracle<String, String> eqOracleW = new WMethodEQOracle.MealyWMethodEQOracle<>(2, mOracleForTesting); case LeeYannakakis: eqOracle = new YannakakisEQOracle<>(alphabets, mOracleForTesting); break;
EquivalenceOracle.MealyEquivalenceOracle<String, String> eqOracleYannakakis = new YannakakisEQOracle<>(alphabets, mOracleForTesting); case WMethod: eqOracle = new WMethodEQOracle.MealyWMethodEQOracle<>(maxDepthForWMethod, mOracleForTesting); break;
case WpMethod: eqOracle = new WpMethodEQOracle.MealyWpMethodEQOracle<>(maxDepthForWMethod, mOracleForTesting); break;
// The chosen oracle to experiment with. }
EquivalenceOracle<MealyMachine<?, String, ?, String>, String, Word<String>> eqOracle = eqOracleYannakakis; assert eqOracle != null;
// Learnlib comes with different learning algorithms System.out.println("Setting up learner");
System.out.println("Setting up learner(s)"); LearningAlgorithm<MealyMachine<?, String, ?, String>, String, Word<String>> learner = null;
ExtensibleLStarMealy<String, String> learnerLStar = new ExtensibleLStarMealy<>(alphabet, mOracleForLearning, Lists.<Word<String>>newArrayList(), ObservationTableCEXHandlers.CLASSIC_LSTAR, ClosingStrategies.CLOSE_SHORTEST); switch (learnAlgorithm){
TTTLearnerMealy<String, String> learnerTTT = new TTTLearnerMealy<>(alphabet, mOracleForLearning, LocalSuffixFinders.FIND_LINEAR); case LStar:
KearnsVaziraniMealy<String, String> learnerKV = new KearnsVaziraniMealy<>(alphabet, mOracleForLearning, false, AcexAnalyzers.BINARY_SEARCH); learner = new ExtensibleLStarMealy<>(alphabet, mOracleForLearning, Lists.<Word<String>>newArrayList(), ObservationTableCEXHandlers.CLASSIC_LSTAR, ClosingStrategies.CLOSE_SHORTEST);
break;
// The chosen learning algorithm case RivestSchapire:
LearningAlgorithm<MealyMachine<?, String, ?, String>, String, Word<String>> learner = learnerLStar; learner = new ExtensibleLStarMealy<>(alphabet, mOracleForLearning, Lists.<Word<String>>newArrayList(), ObservationTableCEXHandlers.RIVEST_SCHAPIRE, ClosingStrategies.CLOSE_SHORTEST);
break;
case TTT:
learner = new TTTLearnerMealy<>(alphabet, mOracleForLearning, LocalSuffixFinders.FIND_LINEAR);
break;
case KearnsVazirani:
learner = new KearnsVaziraniMealy<>(alphabet, mOracleForLearning, false, AcexAnalyzers.BINARY_SEARCH);
break;
}
assert learner != null;
// Here we will perform our experiment. I did not use the Experiment class from LearnLib, as I wanted some // Here we will perform our experiment. I did not use the Experiment class from LearnLib, as I wanted some
@ -106,11 +143,13 @@ public class Main {
learner.startLearning(); learner.startLearning();
while(true) { while(true) {
// String dir = "/Users/joshua/Documents/PhD/Machines/Mealy/esms3/"; if(saveAllHypotheses) {
// String filenameh = dir + "hyp." + stage + ".obf.dot"; String dir = "/Users/joshua/Documents/PhD/Machines/Mealy/esms3/";
// PrintWriter output = new PrintWriter(filenameh); String filename = dir + "hyp." + stage + ".obf.dot";
// GraphDOT.write(learner.getHypothesisModel(), alphabet, output); PrintWriter output = new PrintWriter(filename);
// output.close(); GraphDOT.write(learner.getHypothesisModel(), alphabet, output);
output.close();
}
System.out.println(stage++ + ": " + Calendar.getInstance().getTime()); System.out.println(stage++ + ": " + Calendar.getInstance().getTime());
System.out.println("Hypothesis: " + learner.getHypothesisModel().getStates().size()); System.out.println("Hypothesis: " + learner.getHypothesisModel().getStates().size());