public void Stop_Started_Stopped()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(10000);
Parallel.Invoke(
() => target.Start(),
() =>
{
Thread.Sleep(10);
Assert.AreEqual(GeneticAlgorithmState.Started, target.State);
Assert.IsTrue(target.IsRunning);
target.Stop();
Thread.Sleep(30);
Assert.AreEqual(GeneticAlgorithmState.Stopped, target.State);
Assert.IsFalse(target.IsRunning);
});
Assert.Less(target.Population.Generations.Count, 10000);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 8.8);
}
public void Resume_TerminationReachedAndTerminationNotChanged_Exception()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(10);
target.Start();
Assert.AreEqual(10, target.Population.Generations.Count);
var timeEvolving = target.TimeEvolving.Ticks;
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.Catch <InvalidOperationException>(() =>
{
target.Resume();
}, "Attempt to resume a genetic algorithm with a termination (GenerationNumberTermination (HasReached: True)) already reached. Please, specify a new termination or extend the current one.");
}
public void Start_ParallelManySlowFitness_Timeout()
{
var taskExecutor = new ParallelTaskExecutor();
taskExecutor.MinThreads = 100;
taskExecutor.MaxThreads = 100;
taskExecutor.Timeout = TimeSpan.FromMilliseconds(1000);
var selection = new RouletteWheelSelection();
var crossover = new OnePointCrossover(1);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 150, chromosome),
new FitnessStub()
{
SupportsParallel = true, ParallelSleep = 1500
}, selection, crossover, mutation);
target.TaskExecutor = taskExecutor;
Assert.Catch <TimeoutException>(() =>
{
target.Start();
}, "The fitness evaluation rech the 00:00:01 timeout.");
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(GeneticAlgorithmState.Stopped, target.State);
}
public void Start_TerminationReached_TerminationReachedEventRaised()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Population.GenerationStrategy = new TrackingGenerationStrategy();
target.Termination = new GenerationNumberTermination(1);
var raised = false;
target.TerminationReached += (e, a) =>
{
raised = true;
};
target.Start();
Assert.IsTrue(raised);
}
public void Start_InvalidFitnessEvaluateResult_Exception()
{
var selection = new RouletteWheelSelection();
var crossover = new OnePointCrossover(1);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var fitness = MockRepository.GenerateMock <IFitness>();
fitness.Expect(f => f.Evaluate(null)).IgnoreArguments().Return(1.1);
var target = new GeneticAlgorithm(
new Population(20, 20, chromosome),
fitness, selection, crossover, mutation);
ExceptionAssert.IsThrowing(new FitnessException(fitness, "The {0}.Evaluate returns a fitness with value 1.1. The fitness value should be between 0.0 and 1.0.".With(fitness.GetType())), () => {
target.Start();
});
fitness = MockRepository.GenerateMock <IFitness>();
fitness.Expect(f => f.Evaluate(null)).IgnoreArguments().Return(-0.1);
target = new GeneticAlgorithm(
new Population(20, 20, chromosome),
fitness, selection, crossover, mutation);
ExceptionAssert.IsThrowing(new FitnessException(fitness, "The {0}.Evaluate returns a fitness with value -0.1. The fitness value should be between 0.0 and 1.0.".With(fitness.GetType())), () =>
{
target.Start();
});
}
public void Resume_Stopped_Resumed()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Termination = new TimeEvolvingTermination(TimeSpan.FromMilliseconds(1000));
Parallel.Invoke(
() => target.Start(),
() =>
{
Thread.Sleep(30);
target.Stop();
Thread.Sleep(30);
Assert.AreEqual(GeneticAlgorithmState.Stopped, target.State);
Assert.IsFalse(target.IsRunning);
});
Parallel.Invoke(
() => target.Resume(),
() =>
{
Thread.Sleep(30);
Assert.AreEqual(GeneticAlgorithmState.Resumed, target.State);
Assert.IsTrue(target.IsRunning);
});
}
public void Resume_TerminationReachedAndTerminationExtend_Resumed()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Population.GenerationStrategy = new TrackingGenerationStrategy();
target.Termination = new GenerationNumberTermination(100);
target.Start();
Assert.AreEqual(100, target.Population.Generations.Count);
var timeEvolving = target.TimeEvolving.Ticks;
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(200);
target.Resume();
Assert.AreEqual(target.Population.Generations.Count, 200);
Assert.Less(timeEvolving, target.TimeEvolving.Ticks);
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(300);
target.Resume();
Assert.AreEqual(target.Population.Generations.Count, 300);
Assert.Less(timeEvolving, target.TimeEvolving.Ticks);
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
}
public void Start_ParallelManyGenerations_Optimization()
{
var taskExecutor = new ParallelTaskExecutor();
taskExecutor.MinThreads = 100;
taskExecutor.MaxThreads = 100;
var selection = new EliteSelection();
var crossover = new OnePointCrossover(1);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
FlowAssert.IsAtLeastOneAttemptOk(20, () =>
{
var target = new GeneticAlgorithm(new Population(100, 150, chromosome),
new FitnessStub()
{
SupportsParallel = true
}, selection, crossover, mutation);
target.TaskExecutor = taskExecutor;
Assert.AreEqual(GeneticAlgorithmState.NotStarted, target.State);
Assert.IsFalse(target.IsRunning);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count >= 100);
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count <= 150);
Assert.IsNotNull(target.Population.BestChromosome);
Assert.IsTrue(target.Population.BestChromosome.Fitness >= 0.9);
Assert.IsTrue(target.Population.Generations.Count > 0);
});
FlowAssert.IsAtLeastOneAttemptOk(20, () =>
{
var target = new GeneticAlgorithm(new Population(100, 150, chromosome),
new FitnessStub()
{
SupportsParallel = true
}, selection, crossover, mutation);
target.TaskExecutor = taskExecutor;
target.Start();
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count >= 100);
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count <= 150);
Assert.IsNotNull(target.Population.BestChromosome);
Assert.IsTrue(target.Population.BestChromosome.Fitness >= 0.9);
Assert.IsTrue(target.Population.Generations.Count > 0);
});
}
public void Start_FitnessEvaluationFailed_FitnessException()
{
var selection = new RouletteWheelSelection();
var crossover = new OnePointCrossover(1);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var fitness = new FuncFitness((c) => throw new Exception("TEST"));
var target = new GeneticAlgorithm(new Population(100, 150, chromosome), fitness, selection, crossover, mutation);
Assert.Catch <FitnessException>(target.Start);
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(GeneticAlgorithmState.Stopped, target.State);
}
public void Stop_NotStarted_Exception()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
ExceptionAssert.IsThrowing(new InvalidOperationException("Attempt to stop a genetic algorithm which was not yet started."), () =>
{
target.Stop();
});
}
public void Resume_NotStarted_Exception()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
Assert.Catch <InvalidOperationException>(() =>
{
target.Resume();
}, "Attempt to resume a genetic algorithm which was not yet started.");
}
public void Start_TplGAs_Fast()
{
// GA 1
var selection1 = new EliteSelection();
var crossover1 = new OnePointCrossover(2);
var mutation1 = new UniformMutation();
var chromosome1 = new ChromosomeStub();
var ga1 = new GeneticAlgorithm(new TplPopulation(100, 199, chromosome1),
new FitnessStub()
{
SupportsParallel = false
}, selection1, crossover1, mutation1);
ga1.OperatorsStrategy = new TplOperatorsStrategy();
ga1.Population.GenerationStrategy = new TrackingGenerationStrategy();
ga1.Termination = new GenerationNumberTermination(1000);
// GA 2
var selection2 = new EliteSelection();
var crossover2 = new OnePointCrossover(2);
var mutation2 = new UniformMutation();
var chromosome2 = new ChromosomeStub();
var ga2 = new GeneticAlgorithm(new TplPopulation(100, 199, chromosome2),
new FitnessStub()
{
SupportsParallel = false
}, selection2, crossover2, mutation2);
ga2.OperatorsStrategy = new TplOperatorsStrategy();
ga2.Population.GenerationStrategy = new TrackingGenerationStrategy();
ga2.Termination = new GenerationNumberTermination(1000);
Parallel.Invoke(
() => ga1.Start(),
() => ga2.Start());
Assert.AreEqual(1000, ga1.Population.Generations.Count);
Assert.AreEqual(1000, ga2.Population.Generations.Count);
}
public void Start_ManyCalls_NewEvolutions()
{
RandomizationProvider.Current = new BasicRandomization();
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Population.GenerationStrategy = new TrackingGenerationStrategy();
target.Termination = new GenerationNumberTermination(1000);
Assert.AreEqual(GeneticAlgorithmState.NotStarted, target.State);
Assert.IsFalse(target.IsRunning);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
var lastTimeEvolving = target.TimeEvolving.Ticks;
Assert.AreEqual(1000, target.Population.Generations.Count);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 1);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(1000, target.Population.Generations.Count);
Assert.AreNotEqual(lastTimeEvolving, target.TimeEvolving.Ticks);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(1000, target.Population.Generations.Count);
Assert.AreNotEqual(lastTimeEvolving, target.TimeEvolving.Ticks);
}
public void Start_ParallelManyGenerations_Faster()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 100, chromosome),
new FitnessStub()
{
SupportsParallel = true
}, selection, crossover, mutation);
target.Population.GenerationStrategy = new TrackingGenerationStrategy();
target.Termination = new GenerationNumberTermination(100);
target.TaskExecutor = new ParallelTaskExecutor();
target.Start();
Assert.AreEqual(100, target.Population.Generations.Count);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 1);
}
public void Start_ThreeParentCrossover_KeepsMinSizePopulation()
{
var selection = new EliteSelection();
var crossover = new ThreeParentCrossover();
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(100);
target.Start();
Assert.AreEqual(100, target.Population.Generations.Count);
Assert.IsTrue(target.Population.Generations.All(g => g.Chromosomes.Count >= 100));
}
public void Start_NotParallelManyGenerations_Fast()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(100);
TimeAssert.LessThan(300, () => {
target.Start();
});
Assert.AreEqual(100, target.Population.Generations.Count);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 1);
}
public void Start_NotParallelManyGenerations_Optimization()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(50, 50, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Population.GenerationStrategy = new TrackingGenerationStrategy();
Assert.IsInstanceOf <EliteSelection>(target.Selection);
Assert.IsInstanceOf <OnePointCrossover>(target.Crossover);
Assert.IsInstanceOf <UniformMutation>(target.Mutation);
target.Termination = new GenerationNumberTermination(25);
Assert.AreEqual(GeneticAlgorithmState.NotStarted, target.State);
Assert.IsFalse(target.IsRunning);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(25, target.Population.Generations.Count);
var lastFitness = 0.0;
foreach (var g in target.Population.Generations)
{
Assert.GreaterOrEqual(g.BestChromosome.Fitness.Value, lastFitness);
lastFitness = g.BestChromosome.Fitness.Value;
}
Assert.GreaterOrEqual(lastFitness, 0.8);
Assert.AreEqual(lastFitness, target.BestChromosome.Fitness);
}
public void Start_ManyCallsTerminationChanged_NewEvolutions()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub()
{
SupportsParallel = false
}, selection, crossover, mutation);
target.Population.GenerationStrategy = new TrackingGenerationStrategy();
target.Termination = new GenerationNumberTermination(500);
target.Start();
var lastTimeEvolving = target.TimeEvolving.TotalMilliseconds;
Assert.AreEqual(500, target.Population.Generations.Count);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 1);
Assert.Less(target.TimeEvolving.TotalMilliseconds, 1500, "Time evolving should be less than 1000ms");
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(100);
target.Start();
Assert.AreEqual(100, target.Population.Generations.Count);
Assert.Less(target.TimeEvolving.TotalMilliseconds, lastTimeEvolving, "Time evolving 50 generations should be less than 100-199 generations");
lastTimeEvolving = target.TimeEvolving.TotalMilliseconds;
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(25);
target.Start();
Assert.AreEqual(25, target.Population.Generations.Count);
Assert.Less(target.TimeEvolving.TotalMilliseconds, lastTimeEvolving);
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
}
public void Start_ParallelManyGenerations_Faster()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var fitness = new FitnessStub()
{
SupportsParallel = true, ParallelSleep = 1
};
var generationStrategy = new TrackingGenerationStrategy();
var termination = new GenerationNumberTermination(100);
// we test linear first
var target = new GeneticAlgorithm(new Population(10, 10, chromosome), fitness
, selection, crossover, mutation);
target.Population.GenerationStrategy = generationStrategy;
target.Termination = termination;
target.TaskExecutor = new LinearTaskExecutor();
target.Start();
Assert.AreEqual(100, target.Population.Generations.Count);
var linearTime = target.TimeEvolving.TotalMilliseconds;
//then parallel
target = new GeneticAlgorithm(new Population(10, 10, chromosome), fitness
, selection, crossover, mutation);
target.Population.GenerationStrategy = generationStrategy;
target.Termination = termination;
target.TaskExecutor = new TplTaskExecutor();
target.Start();
Assert.AreEqual(100, target.Population.Generations.Count);
var parallelTime = target.TimeEvolving.TotalMilliseconds;
Assert.Less(parallelTime, linearTime);
}
public void Start_InvalidFitnessEvaluateResult_Exception()
{
var selection = new RouletteWheelSelection();
var crossover = new OnePointCrossover(1);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var fitness = MockRepository.GenerateMock<IFitness>();
fitness.Expect(f => f.Evaluate(null)).IgnoreArguments().Return(1.1);
var target = new GeneticAlgorithm(
new Population(20, 20, chromosome),
fitness, selection, crossover, mutation);
ExceptionAssert.IsThrowing(new FitnessException(fitness, "The {0}.Evaluate returns a fitness with value 1.1. The fitness value should be between 0.0 and 1.0.".With(fitness.GetType())), () =>
{
target.Start();
});
fitness = MockRepository.GenerateMock<IFitness>();
fitness.Expect(f => f.Evaluate(null)).IgnoreArguments().Return(-0.1);
target = new GeneticAlgorithm(
new Population(20, 20, chromosome),
fitness, selection, crossover, mutation);
ExceptionAssert.IsThrowing(new FitnessException(fitness, "The {0}.Evaluate returns a fitness with value -0.1. The fitness value should be between 0.0 and 1.0.".With(fitness.GetType())), () =>
{
target.Start();
});
}
public void Start_ManyCallsTerminationChanged_NewEvolutions()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(100);
target.Start();
var lastTimeEvolving = target.TimeEvolving.TotalMilliseconds;
Assert.AreEqual(100, target.Population.Generations.Count);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 1);
Assert.Less(target.TimeEvolving.TotalMilliseconds, 1000);
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(50);
target.Start();
Assert.AreEqual(50, target.Population.Generations.Count);
Assert.Less(target.TimeEvolving.TotalMilliseconds, lastTimeEvolving);
lastTimeEvolving = target.TimeEvolving.TotalMilliseconds;
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(25);
target.Start();
Assert.AreEqual(25, target.Population.Generations.Count);
Assert.Less(target.TimeEvolving.TotalMilliseconds, lastTimeEvolving);
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
}
public void Start_ParallelManySlowFitness_Timeout()
{
var taskExecutor = new SmartThreadPoolTaskExecutor();
taskExecutor.MinThreads = 100;
taskExecutor.MaxThreads = 100;
taskExecutor.Timeout = TimeSpan.FromMilliseconds(1000);
var selection = new RouletteWheelSelection();
var crossover = new OnePointCrossover(1);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 150, chromosome),
new FitnessStub() { SupportsParallel = true, ParallelSleep = 1500 }, selection, crossover, mutation);
target.TaskExecutor = taskExecutor;
ExceptionAssert.IsThrowing(new TimeoutException("The fitness evaluation rech the 00:00:01 timeout."), () =>
{
target.Start();
});
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(GeneticAlgorithmState.Stopped, target.State);
}
public void Start_ParallelManyGenerations_Optimization()
{
var taskExecutor = new SmartThreadPoolTaskExecutor();
taskExecutor.MinThreads = 100;
taskExecutor.MaxThreads = 100;
var selection = new EliteSelection();
var crossover = new OnePointCrossover(1);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
FlowAssert.IsAtLeastOneAttemptOk(20, () =>
{
var target = new GeneticAlgorithm(new Population(100, 150, chromosome),
new FitnessStub() { SupportsParallel = true }, selection, crossover, mutation);
target.TaskExecutor = taskExecutor;
Assert.AreEqual(GeneticAlgorithmState.NotStarted, target.State);
Assert.IsFalse(target.IsRunning);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count >= 100);
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count <= 150);
Assert.IsNotNull(target.Population.BestChromosome);
Assert.IsTrue(target.Population.BestChromosome.Fitness >= 0.9);
Assert.IsTrue(target.Population.Generations.Count > 0);
});
FlowAssert.IsAtLeastOneAttemptOk(20, () =>
{
var target = new GeneticAlgorithm(new Population(100, 150, chromosome),
new FitnessStub() { SupportsParallel = true }, selection, crossover, mutation);
target.TaskExecutor = taskExecutor;
target.Start();
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count >= 100);
Assert.IsTrue(target.Population.CurrentGeneration.Chromosomes.Count <= 150);
Assert.IsNotNull(target.Population.BestChromosome);
Assert.IsTrue(target.Population.BestChromosome.Fitness >= 0.9);
Assert.IsTrue(target.Population.Generations.Count > 0);
});
}
public void Start_ParallelGAs_Fast()
{
// GA 1
var selection1 = new EliteSelection();
var crossover1 = new OnePointCrossover(2);
var mutation1 = new UniformMutation();
var chromosome1 = new ChromosomeStub();
var ga1 = new GeneticAlgorithm(new Population(100, 199, chromosome1),
new FitnessStub() { SupportsParallel = false }, selection1, crossover1, mutation1);
ga1.Termination = new GenerationNumberTermination(1000);
// GA 2
var selection2 = new EliteSelection();
var crossover2 = new OnePointCrossover(2);
var mutation2 = new UniformMutation();
var chromosome2 = new ChromosomeStub();
var ga2 = new GeneticAlgorithm(new Population(100, 199, chromosome2),
new FitnessStub() { SupportsParallel = false }, selection2, crossover2, mutation2);
ga2.Termination = new GenerationNumberTermination(1000);
Parallel.Invoke(
() => ga1.Start(),
() => ga2.Start());
Assert.AreEqual(1000, ga1.Population.Generations.Count);
Assert.AreEqual(1000, ga2.Population.Generations.Count);
}
public void Start_NotParallelManyGenerations_Fast()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(100);
TimeAssert.LessThan(30000, () =>
{
target.Start();
});
Assert.AreEqual(100, target.Population.Generations.Count);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 1);
}
public void Start_ThreeParentCrossover_KeepsMinSizePopulation()
{
var selection = new EliteSelection();
var crossover = new ThreeParentCrossover();
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(100);
target.Start();
Assert.AreEqual(100, target.Population.Generations.Count);
Assert.IsTrue(target.Population.Generations.All(g => g.Chromosomes.Count >= 100));
}
public void Start_TerminationReached_TerminationReachedEventRaised()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(1);
var raised = false;
target.TerminationReached += (e, a) =>
{
raised = true;
};
target.Start();
Assert.IsTrue(raised);
}
public void Stop_Started_Stopped()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(10000);
Parallel.Invoke(
() => target.Start(),
() =>
{
Thread.Sleep(10);
Assert.AreEqual(GeneticAlgorithmState.Started, target.State);
Assert.IsTrue(target.IsRunning);
target.Stop();
Thread.Sleep(30);
Assert.AreEqual(GeneticAlgorithmState.Stopped, target.State);
Assert.IsFalse(target.IsRunning);
});
Assert.Less(target.Population.Generations.Count, 10000);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 8.8);
}
public void Start_ManyCalls_NewEvolutions()
{
RandomizationProvider.Current = new BasicRandomization();
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(1000);
Assert.AreEqual(GeneticAlgorithmState.NotStarted, target.State);
Assert.IsFalse(target.IsRunning);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
var lastTimeEvolving = target.TimeEvolving.Ticks;
Assert.AreEqual(1000, target.Population.Generations.Count);
Assert.Greater(target.TimeEvolving.TotalMilliseconds, 1);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(1000, target.Population.Generations.Count);
Assert.AreNotEqual(lastTimeEvolving, target.TimeEvolving.Ticks);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(1000, target.Population.Generations.Count);
Assert.AreNotEqual(lastTimeEvolving, target.TimeEvolving.Ticks);
}
public void Start_NotParallelManyGenerations_Optimization()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(50, 50, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
Assert.IsInstanceOf<EliteSelection>(target.Selection);
Assert.IsInstanceOf<OnePointCrossover>(target.Crossover);
Assert.IsInstanceOf<UniformMutation>(target.Mutation);
target.Termination = new GenerationNumberTermination(25);
Assert.AreEqual(GeneticAlgorithmState.NotStarted, target.State);
Assert.IsFalse(target.IsRunning);
target.Start();
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
Assert.AreEqual(25, target.Population.Generations.Count);
var lastFitness = 0.0;
foreach (var g in target.Population.Generations)
{
Assert.GreaterOrEqual(g.BestChromosome.Fitness.Value, lastFitness);
lastFitness = g.BestChromosome.Fitness.Value;
}
Assert.GreaterOrEqual(lastFitness, 0.8);
Assert.AreEqual(lastFitness, target.BestChromosome.Fitness);
}
public void Resume_TerminationReachedAndTerminationExtend_Resumed()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(100);
target.Start();
Assert.AreEqual(target.Population.Generations.Count, 100);
var timeEvolving = target.TimeEvolving.Ticks;
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(200);
target.Resume();
Assert.AreEqual(target.Population.Generations.Count, 200);
Assert.Less(timeEvolving, target.TimeEvolving.Ticks);
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
target.Termination = new GenerationNumberTermination(300);
target.Resume();
Assert.AreEqual(target.Population.Generations.Count, 300);
Assert.Less(timeEvolving, target.TimeEvolving.Ticks);
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
}
public void Resume_TerminationReachedAndTerminationNotChanged_Exception()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new GenerationNumberTermination(10);
target.Start();
Assert.AreEqual(10, target.Population.Generations.Count);
var timeEvolving = target.TimeEvolving.Ticks;
Assert.AreEqual(GeneticAlgorithmState.TerminationReached, target.State);
Assert.IsFalse(target.IsRunning);
ExceptionAssert.IsThrowing(new InvalidOperationException("Attempt to resume a genetic algorithm with a termination (GenerationNumberTermination (HasReached: True)) already reached. Please, specify a new termination or extend the current one."), () =>
{
target.Resume();
});
}
public void Resume_Stopped_Resumed()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
target.Termination = new TimeEvolvingTermination(TimeSpan.FromMilliseconds(10000));
target.TaskExecutor = new SmartThreadPoolTaskExecutor();
var stoppedCount = 0;
target.Stopped += (e, a) =>
{
Assert.AreEqual(GeneticAlgorithmState.Stopped, target.State);
Assert.IsFalse(target.IsRunning);
stoppedCount++;
};
Parallel.Invoke(
() => target.Start(),
() =>
{
Thread.Sleep(500);
target.Stop();
});
Thread.Sleep(2000);
Parallel.Invoke(
() => target.Resume(),
() =>
{
Thread.Sleep(2000);
Assert.AreEqual(GeneticAlgorithmState.Resumed, target.State);
Assert.IsTrue(target.IsRunning);
});
Assert.AreEqual(1, stoppedCount);
}
public void Resume_NotStarted_Exception()
{
var selection = new EliteSelection();
var crossover = new OnePointCrossover(2);
var mutation = new UniformMutation();
var chromosome = new ChromosomeStub();
var target = new GeneticAlgorithm(new Population(100, 199, chromosome),
new FitnessStub() { SupportsParallel = false }, selection, crossover, mutation);
ExceptionAssert.IsThrowing(new InvalidOperationException("Attempt to resume a genetic algorithm which was not yet started."), () =>
{
target.Resume();
});
}
