public void Descriptor_Save_And_Load_Json()
{
var data = Iris.Load();
var description = Descriptor.Create <Iris>();
// to populate dictionaries
var examples = description.ToExamples(data);
var file = GetPath();
Register.Type <Iris>();
var d = SaveAndLoadJson(description);
Assert.Equal(description.Type, d.Type);
Assert.Equal(description.Features.Length, d.Features.Length);
for (int i = 0; i < description.Features.Length; i++)
{
Assert.Equal(description.Features[i].Type, d.Features[i].Type);
Assert.Equal(description.Features[i].Name, d.Features[i].Name);
Assert.Equal(description.Features[i].Start, d.Features[i].Start);
}
Assert.Equal(description.Label.Type, d.Label.Type);
Assert.Equal(description.Label.Name, d.Label.Name);
Assert.Equal(description.Label.Start, d.Label.Start);
}
public void Iris_Naive_Bayes_Save_And_Load_Test()
{
var data = Iris.Load();
var description = Descriptor.Create <Iris>();
var generator = new NaiveBayesGenerator(2);
var model = generator.Generate(description, data);
Serialize(model);
var lmodel = Deserialize <NaiveBayesModel>();
}
public void Save_And_Load_Iris_DT()
{
var data = Iris.Load();
var description = Descriptor.Create <Iris>();
var generator = new DecisionTreeGenerator(50);
var model = generator.Generate(description, data) as DecisionTreeModel;
Serialize(model);
var lmodel = Deserialize <DecisionTreeModel>();
Assert.AreEqual(model.Hint, lmodel.Hint);
AreEqual(model.Tree, lmodel.Tree, false);
}
public void Iris_Naive_Bayes_Save_And_Load_Test_Json()
{
var data = Iris.Load();
var description = Descriptor.Create <Iris>();
var generator = new NaiveBayesGenerator(2);
var model = generator.Generate(description, data) as NaiveBayesModel;
var file = GetPath();
Register.Type <Iris>();
var lmodel = SaveAndLoadJson(model);
Assert.Equal(model.Root, lmodel.Root);
}
public void Save_And_Load_Iris_DT_Json()
{
var data = Iris.Load();
var description = Descriptor.Create <Iris>();
var generator = new DecisionTreeGenerator(50);
var model = generator.Generate(description, data) as DecisionTreeModel;
var file = GetPath();
Register.Type <Iris>();
var lmodel = SaveAndLoadJson(model);
Assert.Equal(model.Descriptor, lmodel.Descriptor);
Assert.Equal(model.Hint, lmodel.Hint);
Assert.Equal(model.Tree, lmodel.Tree);
}
public void Iris_Tests()
{
// need to run multiple times since
// this model is a bit more sensitive
LearnerPrediction <Iris>(
new NeuralNetworkGenerator(),
Iris.Load(),
new Iris
{
PetalWidth = 0.5m,
PetalLength = 2.3m,
SepalLength = 2.1m,
SepalWidth = 2.1m
},
i => "Iris-setosa".Sanitize() == i.Class
);
}
public void IrisPrediction(IGenerator generator)
{
// should be Iris-Setosa
Iris iris = new Iris
{
PetalWidth = 0.5m,
PetalLength = 2.3m,
SepalLength = 2.1m,
SepalWidth = 2.1m
};
Prediction <Iris>(
generator, // DT
Iris.Load(), // training data
iris, // test object
i => "Iris-setosa".Sanitize() == i.Class // should be true
);
}
public void IrisLearnerPrediction(IGenerator generator)
{
// should be Iris-Setosa
Iris iris = new Iris
{
PetalWidth = 0.5m,
PetalLength = 2.3m,
SepalLength = 2.1m,
SepalWidth = 2.1m
};
LearnerPrediction <Iris>(
generator,
Iris.Load(),
iris,
i => "Iris-setosa".Sanitize() == i.Class // should be true
);
}
public void Iris_DT_and_Prediction()
{
var data = Iris.Load();
var description = Descriptor.Create <Iris>();
var generator = new DecisionTreeGenerator(50);
var model = generator.Generate(description, data);
// should be Iris-Setosa
Iris iris = new Iris
{
PetalWidth = 0.5m,
PetalLength = 2.3m,
SepalLength = 2.1m,
SepalWidth = 2.1m
};
iris = model.Predict <Iris>(iris);
Assert.AreEqual("Iris-setosa".Sanitize(), iris.Class);
}
public void Descriptor_Save_And_load()
{
var data = Iris.Load();
var description = Descriptor.Create <Iris>();
// to populate dictionaries
var examples = description.ToExamples(data);
Serialize(description);
var d = Deserialize <Descriptor>();
Assert.Equal(description.Type, d.Type);
Assert.Equal(description.Features.Length, d.Features.Length);
for (int i = 0; i < description.Features.Length; i++)
{
Assert.Equal(description.Features[i].Type, d.Features[i].Type);
Assert.Equal(description.Features[i].Name, d.Features[i].Name);
Assert.Equal(description.Features[i].Start, d.Features[i].Start);
}
Assert.Equal(description.Label.Type, d.Label.Type);
Assert.Equal(description.Label.Name, d.Label.Name);
Assert.Equal(description.Label.Start, d.Label.Start);
}
public void Iris_DT_and_Prediction_with_Learner()
{
var data = Iris.Load();
var generator = new DecisionTreeGenerator(50)
{
Descriptor = Descriptor.Create <Iris>(),
Hint = 0
};
var lmodel = Learner.Learn(data, .80, 1000, generator);
// should be Iris-Setosa
Iris iris = new Iris
{
PetalWidth = 0.5m,
PetalLength = 2.3m,
SepalLength = 2.1m,
SepalWidth = 2.1m
};
iris = lmodel.Model.Predict <Iris>(iris);
Assert.AreEqual("Iris-setosa".Sanitize(), iris.Class);
}
public void Iris_FluentDescriptor_DT_and_Prediction()
{
var data = Iris.Load();
var description = Descriptor.For <Iris>()
.Learn(x => x.Class)
.With(x => x.PetalLength)
.With(x => x.PetalWidth)
.With(x => x.SepalLength)
.With(x => x.SepalWidth);
var generator = new NaiveBayesGenerator(2);
var model = generator.Generate(description, data);
// should be Iris-Setosa
Iris iris = new Iris
{
SepalLength = 2.1m,
SepalWidth = 2.2m,
PetalWidth = 0.5m,
PetalLength = 2.3m,
};
iris = model.Predict <Iris>(iris);
Assert.AreEqual("Iris-setosa".Sanitize(), iris.Class);
}
