6673301747

Committed 27 Oct 2023 11:46PM UTC coverage: 91.986% (-1.3%) from 93.327%

Build # 6673301747

Build Type

push

github

Committed by

mtf90

Commit Message

merge the release and sign-artifacts profiles

Run Details

10984 of 11941 relevant lines covered (91.99%)

1.72 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

86.44

/commons/util/src/main/java/de/learnlib/util/Experiment.java

/* Copyright (C) 2013-2023 TU Dortmund
 * This file is part of LearnLib, http://www.learnlib.de/.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
package de.learnlib.util;

import de.learnlib.api.algorithm.LearningAlgorithm;
import de.learnlib.api.logging.Category;
import de.learnlib.api.oracle.EquivalenceOracle;
import de.learnlib.api.query.DefaultQuery;
import de.learnlib.filter.statistic.Counter;
import de.learnlib.util.statistic.SimpleProfiler;
import net.automatalib.alphabet.Alphabet;
import net.automatalib.automaton.fsa.DFA;
import net.automatalib.automaton.transducer.MealyMachine;
import net.automatalib.automaton.transducer.MooreMachine;
import net.automatalib.word.Word;
import org.checkerframework.checker.nullness.qual.Nullable;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

/**
 * runs a learning experiment.
 *
 * @param <A> the automaton type
 */
public class Experiment<A extends Object> {

    public static final String LEARNING_PROFILE_KEY = "Learning";
    public static final String COUNTEREXAMPLE_PROFILE_KEY = "Searching for counterexample";

    private static final Logger LOGGER = LoggerFactory.getLogger(Experiment.class);
    private final ExperimentImpl<?, ?> impl;
    private boolean logModels;
    private boolean profile;
    private final Counter rounds = new Counter("Learning rounds", "#");
    private @Nullable A finalHypothesis;

    public <I, D> Experiment(LearningAlgorithm<? extends A, I, D> learningAlgorithm,
                             EquivalenceOracle<? super A, I, D> equivalenceAlgorithm,
                             Alphabet<I> inputs) {
        this.impl = new ExperimentImpl<>(learningAlgorithm, equivalenceAlgorithm, inputs);
    }

    public A run() {
        if (this.finalHypothesis != null) {
            throw new IllegalStateException("Experiment has already been run");
        }

        finalHypothesis = impl.run();
        return finalHypothesis;
    }

    public A getFinalHypothesis() {
        if (finalHypothesis == null) {
            throw new IllegalStateException("Experiment has not yet been run");
        }

        return finalHypothesis;
    }

    private void profileStart(String taskname) {
        if (profile) {
            SimpleProfiler.start(taskname);
        }
    }

    private void profileStop(String taskname) {
        if (profile) {
            SimpleProfiler.stop(taskname);
        }
    }

    /**
     * @param logModels
     *         flag whether models should be logged
     */
    public void setLogModels(boolean logModels) {
        this.logModels = logModels;
    }

    /**
     * @param profile
     *         flag whether learning process should be profiled
     */
    public void setProfile(boolean profile) {
        this.profile = profile;
    }

    /**
     * @return the rounds
     */
    public Counter getRounds() {
        return rounds;
    }

    private final class ExperimentImpl<I, D> {

        private final LearningAlgorithm<? extends A, I, D> learningAlgorithm;
        private final EquivalenceOracle<? super A, I, D> equivalenceAlgorithm;
        private final Alphabet<I> inputs;

        ExperimentImpl(LearningAlgorithm<? extends A, I, D> learningAlgorithm,
                       EquivalenceOracle<? super A, I, D> equivalenceAlgorithm,
                       Alphabet<I> inputs) {
            this.learningAlgorithm = learningAlgorithm;
            this.equivalenceAlgorithm = equivalenceAlgorithm;
            this.inputs = inputs;
        }

        public A run() {
            rounds.increment();
            LOGGER.info(Category.PHASE, "Starting round {}", rounds.getCount());
            LOGGER.info(Category.PHASE, "Learning");

            profileStart(LEARNING_PROFILE_KEY);
            learningAlgorithm.startLearning();
            profileStop(LEARNING_PROFILE_KEY);

            while (true) {
                final A hyp = learningAlgorithm.getHypothesisModel();

                if (logModels) {
                    LOGGER.info(Category.MODEL, hyp.toString());
                }

                LOGGER.info(Category.PHASE, "Searching for counterexample");

                profileStart(COUNTEREXAMPLE_PROFILE_KEY);
                DefaultQuery<I, D> ce = equivalenceAlgorithm.findCounterExample(hyp, inputs);
                profileStop(COUNTEREXAMPLE_PROFILE_KEY);

                if (ce == null) {
                    return hyp;
                }

                LOGGER.info(Category.COUNTEREXAMPLE, ce.getInput().toString());

                // next round ...
                rounds.increment();
                LOGGER.info(Category.PHASE, "Starting round {}", rounds.getCount());
                LOGGER.info(Category.PHASE, "Learning");

                profileStart(LEARNING_PROFILE_KEY);
                final boolean refined = learningAlgorithm.refineHypothesis(ce);
                profileStop(LEARNING_PROFILE_KEY);

                assert refined;
            }
        }
    }

    public static class DFAExperiment<I> extends Experiment<DFA<?, I>> {
        public DFAExperiment(LearningAlgorithm<? extends DFA<?, I>, I, Boolean> learningAlgorithm,
                             EquivalenceOracle<? super DFA<?, I>, I, Boolean> equivalenceAlgorithm,
                             Alphabet<I> inputs) {
            super(learningAlgorithm, equivalenceAlgorithm, inputs);
        }

    }

    public static class MealyExperiment<I, O> extends Experiment<MealyMachine<?, I, ?, O>> {

        public MealyExperiment(LearningAlgorithm<? extends MealyMachine<?, I, ?, O>, I, Word<O>> learningAlgorithm,
                               EquivalenceOracle<? super MealyMachine<?, I, ?, O>, I, Word<O>> equivalenceAlgorithm,
                               Alphabet<I> inputs) {
            super(learningAlgorithm, equivalenceAlgorithm, inputs);
        }

    }

    public static class MooreExperiment<I, O> extends Experiment<MooreMachine<?, I, ?, O>> {

        public MooreExperiment(LearningAlgorithm<? extends MooreMachine<?, I, ?, O>, I, Word<O>> learningAlgorithm,
                               EquivalenceOracle<? super MooreMachine<?, I, ?, O>, I, Word<O>> equivalenceAlgorithm,
                               Alphabet<I> inputs) {
            super(learningAlgorithm, equivalenceAlgorithm, inputs);
        }

    }

}

1	/* Copyright (C) 2013-2023 TU Dortmund
2	* This file is part of LearnLib, http://www.learnlib.de/.
3	*
4	* Licensed under the Apache License, Version 2.0 (the "License");
5	* you may not use this file except in compliance with the License.
6	* You may obtain a copy of the License at
7	*
8	* http://www.apache.org/licenses/LICENSE-2.0
9	*
10	* Unless required by applicable law or agreed to in writing, software
11	* distributed under the License is distributed on an "AS IS" BASIS,
12	* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13	* See the License for the specific language governing permissions and
14	* limitations under the License.
15	*/
16	package de.learnlib.util;
17
18	import de.learnlib.api.algorithm.LearningAlgorithm;
19	import de.learnlib.api.logging.Category;
20	import de.learnlib.api.oracle.EquivalenceOracle;
21	import de.learnlib.api.query.DefaultQuery;
22	import de.learnlib.filter.statistic.Counter;
23	import de.learnlib.util.statistic.SimpleProfiler;
24	import net.automatalib.alphabet.Alphabet;
25	import net.automatalib.automaton.fsa.DFA;
26	import net.automatalib.automaton.transducer.MealyMachine;
27	import net.automatalib.automaton.transducer.MooreMachine;
28	import net.automatalib.word.Word;
29	import org.checkerframework.checker.nullness.qual.Nullable;
30	import org.slf4j.Logger;
31	import org.slf4j.LoggerFactory;
32
33	/**
34	* runs a learning experiment.
35	*
36	* @param <A> the automaton type
37	*/
38	public class Experiment<A extends Object> {
39
40	public static final String LEARNING_PROFILE_KEY = "Learning";
41	public static final String COUNTEREXAMPLE_PROFILE_KEY = "Searching for counterexample";
42
43	private static final Logger LOGGER = LoggerFactory.getLogger(Experiment.class);	2✔
44	private final ExperimentImpl<?, ?> impl;
45	private boolean logModels;
46	private boolean profile;
47	private final Counter rounds = new Counter("Learning rounds", "#");	2✔
48	private @Nullable A finalHypothesis;
49
50	public <I, D> Experiment(LearningAlgorithm<? extends A, I, D> learningAlgorithm,
51	EquivalenceOracle<? super A, I, D> equivalenceAlgorithm,
52	Alphabet<I> inputs) {	2✔
53	this.impl = new ExperimentImpl<>(learningAlgorithm, equivalenceAlgorithm, inputs);	2✔
54	}	2✔
55
56	public A run() {
57	if (this.finalHypothesis != null) {	2✔
58	throw new IllegalStateException("Experiment has already been run");	2✔
59	}
60
61	finalHypothesis = impl.run();	2✔
62	return finalHypothesis;	2✔
63	}
64
65	public A getFinalHypothesis() {
66	if (finalHypothesis == null) {	2✔
67	throw new IllegalStateException("Experiment has not yet been run");	2✔
68	}
69
70	return finalHypothesis;	2✔
71	}
72
73	private void profileStart(String taskname) {
74	if (profile) {	2✔
75	SimpleProfiler.start(taskname);	2✔
76	}
77	}	2✔
78
79	private void profileStop(String taskname) {
80	if (profile) {	2✔
81	SimpleProfiler.stop(taskname);	2✔
82	}
83	}	2✔
84
85	/**
86	* @param logModels
87	* flag whether models should be logged
88	*/
89	public void setLogModels(boolean logModels) {
90	this.logModels = logModels;	×
91	}	×
92
93	/**
94	* @param profile
95	* flag whether learning process should be profiled
96	*/
97	public void setProfile(boolean profile) {
98	this.profile = profile;	2✔
99	}	2✔
100
101	/**
102	* @return the rounds
103	*/
104	public Counter getRounds() {
105	return rounds;	×
106	}
107
108	private final class ExperimentImpl<I, D> {	2✔
109
110	private final LearningAlgorithm<? extends A, I, D> learningAlgorithm;
111	private final EquivalenceOracle<? super A, I, D> equivalenceAlgorithm;
112	private final Alphabet<I> inputs;
113
114	ExperimentImpl(LearningAlgorithm<? extends A, I, D> learningAlgorithm,
115	EquivalenceOracle<? super A, I, D> equivalenceAlgorithm,
116	Alphabet<I> inputs) {	2✔
117	this.learningAlgorithm = learningAlgorithm;	2✔
118	this.equivalenceAlgorithm = equivalenceAlgorithm;	2✔
119	this.inputs = inputs;	2✔
120	}	2✔
121
122	public A run() {
123	rounds.increment();	2✔
124	LOGGER.info(Category.PHASE, "Starting round {}", rounds.getCount());	2✔
125	LOGGER.info(Category.PHASE, "Learning");	2✔
126
127	profileStart(LEARNING_PROFILE_KEY);	2✔
128	learningAlgorithm.startLearning();	2✔
129	profileStop(LEARNING_PROFILE_KEY);	2✔
130
131	while (true) {
132	final A hyp = learningAlgorithm.getHypothesisModel();	2✔
133
134	if (logModels) {	2✔
135	LOGGER.info(Category.MODEL, hyp.toString());	×
136	}
137
138	LOGGER.info(Category.PHASE, "Searching for counterexample");	2✔
139
140	profileStart(COUNTEREXAMPLE_PROFILE_KEY);	2✔
141	DefaultQuery<I, D> ce = equivalenceAlgorithm.findCounterExample(hyp, inputs);	2✔
142	profileStop(COUNTEREXAMPLE_PROFILE_KEY);	2✔
143
144	if (ce == null) {	2✔
145	return hyp;	2✔
146	}
147
148	LOGGER.info(Category.COUNTEREXAMPLE, ce.getInput().toString());	2✔
149
150	// next round ...
151	rounds.increment();	2✔
152	LOGGER.info(Category.PHASE, "Starting round {}", rounds.getCount());	2✔
153	LOGGER.info(Category.PHASE, "Learning");	2✔
154
155	profileStart(LEARNING_PROFILE_KEY);	2✔
156	final boolean refined = learningAlgorithm.refineHypothesis(ce);	2✔
157	profileStop(LEARNING_PROFILE_KEY);	2✔
158
159	assert refined;	2✔
160	}	2✔
161	}
162	}
163
164	public static class DFAExperiment<I> extends Experiment<DFA<?, I>> {
165	public DFAExperiment(LearningAlgorithm<? extends DFA<?, I>, I, Boolean> learningAlgorithm,
166	EquivalenceOracle<? super DFA<?, I>, I, Boolean> equivalenceAlgorithm,
167	Alphabet<I> inputs) {
168	super(learningAlgorithm, equivalenceAlgorithm, inputs);	2✔
169	}	2✔
170
171	}
172
173	public static class MealyExperiment<I, O> extends Experiment<MealyMachine<?, I, ?, O>> {
174
175	public MealyExperiment(LearningAlgorithm<? extends MealyMachine<?, I, ?, O>, I, Word<O>> learningAlgorithm,
176	EquivalenceOracle<? super MealyMachine<?, I, ?, O>, I, Word<O>> equivalenceAlgorithm,
177	Alphabet<I> inputs) {
178	super(learningAlgorithm, equivalenceAlgorithm, inputs);	×
179	}	×
180
181	}
182
183	public static class MooreExperiment<I, O> extends Experiment<MooreMachine<?, I, ?, O>> {
184
185	public MooreExperiment(LearningAlgorithm<? extends MooreMachine<?, I, ?, O>, I, Word<O>> learningAlgorithm,
186	EquivalenceOracle<? super MooreMachine<?, I, ?, O>, I, Word<O>> equivalenceAlgorithm,
187	Alphabet<I> inputs) {
188	super(learningAlgorithm, equivalenceAlgorithm, inputs);	×
189	}	×
190
191	}
192
193	}

LearnLib / learnlib / 6673301747

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous