public void TestConstrainedFeasibleRegion()
{
// Test that the parameter set does not deal with its constraints,
// because the URS should, to try to preserve uniformity of sampling with varying bounds.
var p = new TwoParamsConstraints();
Assert.IsTrue(p.IsWithinBounds);
var names = p.GetVariableNames();
Assert.AreEqual(5, p.GetValue(names[1]));
Assert.AreEqual(5, p.GetMaxValue(names[1]));
p.SetValue(names[1], 10);
Assert.IsFalse(p.IsWithinBounds);
Assert.AreEqual(10, p.GetValue(names[1]));
var rng = new BasicRngFactory(0);
var urs = new UniformRandomSamplingFactory <TwoParamsConstraints>(rng.CreateFactory(), new TwoParamsConstraints());
int n = 200;
var pop = new TwoParamsConstraints[n];
for (int i = 0; i < n; i++)
{
pop[i] = urs.CreateRandomCandidate();
Assert.IsTrue(pop[i].IsWithinBounds);
}
var hcOps = urs.CreateNew(rng.CreateFactory());
var hcArray = Array.ConvertAll(pop, x => (IHyperCube <double>)x);
for (int i = 0; i < n; i++)
{
p = (TwoParamsConstraints)hcOps.GenerateRandomWithinHypercube(hcArray);
Assert.IsTrue(p.IsWithinBounds);
}
}
public void TestCoeffVariationTerminationCriteria()
{
double threshold = 2.5e-2;
var termination = new ShuffledComplexEvolution <TestHyperCube> .CoefficientOfVariationTerminationCondition(threshold : threshold);
IObjectiveScores[] population = createSample(converged: false);
double cv = termination.GetMaxParameterCoeffVar(population);
Assert.AreEqual(0.1104401, cv, 1e-6);
Assert.IsFalse(termination.IsBelowCvThreshold(population));
population = createSample(converged: true);
cv = termination.GetMaxParameterCoeffVar(population);
Assert.AreEqual(0.01741291, cv, 1e-6);
Assert.IsTrue(termination.IsBelowCvThreshold(population));
termination = new ShuffledComplexEvolution <TestHyperCube> .CoefficientOfVariationTerminationCondition(threshold : threshold, maxHours : 0.1);
var rng = new BasicRngFactory(0);
var evaluator = new ObjEvalTestHyperCube(new ParaboloidObjEval <TestHyperCube>(bestParam: 2));
var engine = createSce(termination, rng, evaluator);
var results = engine.Evolve();
//Console.WriteLine("Current shuffle: {0}", engine.CurrentShuffle);
Assert.IsFalse(termination.HasReachedMaxTime());
}
private ICandidateFactory <MpiSysConfig> CreatePopulationInitialiser(
IClonableObjectiveEvaluator <MpiSysConfig> objectiveEvaluator,
MpiSysConfig templateParameterSet,
MpiSysConfig[] seedsPopulation,
string initialisationOption = "")
{
Log.Debug("Root: creating population initialiser");
var options = initialisationOption.Split(':');
var seeds = new List <MpiSysConfig> {
templateParameterSet
};
if (seedsPopulation != null)
{
seeds.AddRange(seedsPopulation);
}
var rng = new BasicRngFactory(SceParameterDefinition.RandomizationSeed);
ICandidateFactory <MpiSysConfig> result = CreatePopulationInitialiser(
templateParameterSet,
rng,
objectiveEvaluator,
options);
if (result is BestOfSampling <MpiSysConfig> )
{
return(result);
}
else
{
return(new SeededSamplingFactory <MpiSysConfig>(result, seeds));
}
}
private static ShuffledComplexEvolution <ICloneableSystemConfiguration> createSce(IClonableObjectiveEvaluator <ICloneableSystemConfiguration> evaluator)
{
var rng = new BasicRngFactory(0);
var engine = new ShuffledComplexEvolution <ICloneableSystemConfiguration>(
evaluator,
new UniformRandomSamplingFactory <IHyperCube <double> >(rng.CreateFactory(), new UnivariateReal(0)),
new ShuffledComplexEvolution <ICloneableSystemConfiguration> .MaxShuffleTerminationCondition(),
5, 20, 10, 3, 20, 7,
rng,
new ZitlerThieleFitnessAssignment());
return(engine);
}
private static ICandidateFactory <MpiSysConfig> CreatePopulationInitialiser(
MpiSysConfig templateParameterSet,
BasicRngFactory rng,
IClonableObjectiveEvaluator <MpiSysConfig> objectiveEvaluator,
params string[] options)
{
switch (options[0].ToLower())
{
case "bestof":
int poolSize = int.Parse(options[1]);
int bestPoints = int.Parse(options[2]);
string[] subOptions = Subset(3, options);
if (poolSize < bestPoints)
{
Log.WarnFormat(
"Root: Best of Sampling pool size ({0}) is smaller than best points ({1}). Setting best points to match pool size.",
poolSize,
bestPoints);
bestPoints = poolSize;
}
Log.DebugFormat("Root: creating BestOfSampling. Pool size: {0}, best points: {1}", poolSize, bestPoints);
return
(new BestOfSampling <MpiSysConfig>(
CreatePopulationInitialiser(templateParameterSet, rng, objectiveEvaluator, subOptions),
poolSize,
bestPoints,
objectiveEvaluator,
new DefaultFitnessAssignment()));
case "lhc":
Log.Debug("Root: creating Latin Hypercube Sampling");
return(new LatinHypercubeSampling <MpiSysConfig>(rng, templateParameterSet));
case "weibull":
Log.Debug("Root: creating weibull distribution");
return(new SeededSamplingFactory <MpiSysConfig>(
new WeibullGen <MpiSysConfig>(rng, templateParameterSet),
new List <MpiSysConfig> {
templateParameterSet
}));
case "":
case "urs":
default:
Log.Debug("Root: creating URS");
return(new UniformRandomSamplingFactory <MpiSysConfig>(rng, templateParameterSet));
}
}
private static ShuffledComplexEvolution <TestHyperCube> createSce(double cvThreshold, double maxHours, double objCalcPauseSec, TestObjEval <TestHyperCube> innerObjCalc = null, int maxShuffle = 15)
{
var termination = new ShuffledComplexEvolution <TestHyperCube> .CoefficientOfVariationTerminationCondition(threshold : cvThreshold, maxHours : maxHours);
var rng = new BasicRngFactory(0);
if (innerObjCalc == null)
{
innerObjCalc = new ParaboloidObjEval <TestHyperCube>(bestParam: 2);
}
var evaluator = new ObjEvalTestHyperCube(innerObjCalc, pauseSeconds: objCalcPauseSec);
var engine = createSce(termination, rng, evaluator, maxShuffle: maxShuffle);
return(engine);
}
private static ShuffledComplexEvolution <TestHyperCube> createSce(ShuffledComplexEvolution <TestHyperCube> .CoefficientOfVariationTerminationCondition termination, BasicRngFactory rng, ObjEvalTestHyperCube evaluator, int maxShuffle = 15)
{
var engine = new ShuffledComplexEvolution <TestHyperCube>(
evaluator,
new UniformRandomSamplingFactory <TestHyperCube>(rng.CreateFactory(), new TestHyperCube(2, 0, -10, 10)),
termination,
5, 20, 10, 3, 20, maxShuffle,
rng,
new DefaultFitnessAssignment());
return(engine);
}
