internal void Evaluate_WithVariousValidValues_ReturnsExpectedResult(
double membershipA,
double membershipB,
double expected)
{
// Arrange
var minimumTNorm = TriangularNormFactory.MinimumTNorm();
// Act
var result = minimumTNorm.Evaluate(UnitInterval.Create(membershipA), UnitInterval.Create(membershipB));
// Assert
Assert.Equal(UnitInterval.Create(expected), result);
}
internal void Or_WhenFunctionsDefault_ReturnsExpectedResult()
{
// Arrange
var tNorm = TriangularNormFactory.MinimumTNorm();
var tConorm = TriangularConormFactory.MaximumTConorm();
var evaluator = new FuzzyEvaluator(tNorm, tConorm);
var membershipA = UnitInterval.Create(0.25);
var membershipB = UnitInterval.Create(0.5);
// Act
var result = evaluator.Or(membershipA, membershipB);
// Assert
Assert.Equal(UnitInterval.Create(0.5), result);
}
/// <summary>
/// Initializes a new instance of the <see cref="FuzzyEvaluator"/> class.
/// </summary>
public FuzzyEvaluator()
{
this.triangularNorm = TriangularNormFactory.MinimumTNorm();
this.triangularConorm = TriangularConormFactory.MaximumTConorm();
}
internal void RunMamdaniInference(double foodInput, double serviceInput, double expected)
{
// Define the input and output linguistic variables.
var foodQuality = new LinguisticVariable(
InputVariable.FoodQuality,
new List <FuzzySet>
{
new FuzzySet(FoodQuality.Poor, TrapezoidalFunction.CreateWithLeftEdge(0, 5)),
new FuzzySet(FoodQuality.Average, TriangularFunction.Create(0, 5, 10)),
new FuzzySet(FoodQuality.Good, TrapezoidalFunction.CreateWithRightEdge(5, 10))
});
var serviceQuality = new LinguisticVariable(
InputVariable.FoodQuality,
new List <FuzzySet>
{
new FuzzySet(ServiceQuality.Poor, TrapezoidalFunction.CreateWithLeftEdge(0, 5)),
new FuzzySet(ServiceQuality.Average, TriangularFunction.Create(0, 5, 10)),
new FuzzySet(ServiceQuality.Good, TrapezoidalFunction.CreateWithRightEdge(5, 10))
});
var tipAmount = new LinguisticVariable(
OutputVariable.TipAmount,
new List <FuzzySet>
{
new FuzzySet(TipAmount.Low, TrapezoidalFunction.CreateWithLeftEdge(0, 13)),
new FuzzySet(TipAmount.Medium, TriangularFunction.Create(0, 13, 25)),
new FuzzySet(TipAmount.High, TrapezoidalFunction.CreateWithRightEdge(13, 25))
});
// Define the rules for the fuzzy inference engine.
var rule1 = new FuzzyRuleBuilder(TippingProblem.Rule1)
.If(foodQuality.Is(FoodQuality.Poor))
.Or(serviceQuality.Is(ServiceQuality.Poor))
.Then(tipAmount.Is(TipAmount.Low))
.Build();
var rule2 = new FuzzyRuleBuilder(TippingProblem.Rule2)
.If(serviceQuality.Is(ServiceQuality.Average))
.Then(tipAmount.Is(TipAmount.Medium))
.Build();
var rule3 = new FuzzyRuleBuilder(TippingProblem.Rule3)
.If(foodQuality.Is(FoodQuality.Good))
.Or(serviceQuality.Is(ServiceQuality.Good))
.Then(tipAmount.Is(TipAmount.High))
.Build();
// Construct the fuzzy inference engine.
var tnorm = TriangularNormFactory.MinimumTNorm();
var tconorm = TriangularConormFactory.MaximumTConorm();
var defuzzifier = new CentroidDefuzzifier();
var fuzzyEngine = new MamdaniInferenceEngine(tnorm, tconorm, defuzzifier);
// Add the rules to the rulebase.
fuzzyEngine.Rulebase.AddRule(rule1);
fuzzyEngine.Rulebase.AddRule(rule2);
fuzzyEngine.Rulebase.AddRule(rule3);
// Prepare database to receive inputs.
fuzzyEngine.Database.AddVariable(Label.Create(InputVariable.FoodQuality));
fuzzyEngine.Database.AddVariable(Label.Create(InputVariable.ServiceQuality));
// Generate input data.
var foodData = new DataPoint(InputVariable.FoodQuality, foodInput);
var serviceData = new DataPoint(InputVariable.ServiceQuality, serviceInput);
// Feed inference engine the data.
fuzzyEngine.Database.UpdateData(foodData);
fuzzyEngine.Database.UpdateData(serviceData);
// Compute the inference engine.
var result = fuzzyEngine.Compute();
Assert.Equal(OutputVariable.TipAmount.ToString(), result[0].Subject.Value);
Assert.Equal(expected, result[0].Value);
}