public void F64MatrixExtensions_Map2()
{
var matrix = new F64Matrix(InputData.ToArray(), 2, 3);
matrix.Map(() => 10);
matrix.Map(v => v + 1);
var expected = Enumerable.Range(0, matrix.Data().Length).Select(v => 11.0).ToArray();
CollectionAssert.AreEqual(expected, matrix.Data());
}
public void RegressionNeuralNetModel_Save()
{
var numberOfObservations = 500;
var numberOfFeatures = 5;
var random = new Random(32);
var observations = new F64Matrix(numberOfObservations, numberOfFeatures);
observations.Map(() => random.NextDouble());
var targets = Enumerable.Range(0, numberOfObservations)
.Select(i => (double)random.NextDouble()).ToArray();
var net = new NeuralNet();
net.Add(new InputLayer(numberOfFeatures));
net.Add(new DenseLayer(10));
net.Add(new SquaredErrorRegressionLayer());
var learner = new RegressionNeuralNetLearner(net, new AccuracyLoss());
var sut = learner.Learn(observations, targets);
// save model.
var writer = new StringWriter();
sut.Save(() => writer);
// load model and assert prediction results.
sut = RegressionNeuralNetModel.Load(() => new StringReader(writer.ToString()));
var predictions = sut.Predict(observations);
var evaluator = new MeanSquaredErrorRegressionMetric();
var actual = evaluator.Error(targets, predictions);
Assert.AreEqual(0.0871547675024143, actual, 0.0001);
}
public void RegressionNeuralNetModel_Predict_Single()
{
var numberOfObservations = 500;
var numberOfFeatures = 5;
var random = new Random(32);
var observations = new F64Matrix(numberOfObservations, numberOfFeatures);
observations.Map(() => random.NextDouble());
var targets = Enumerable.Range(0, numberOfObservations)
.Select(i => (double)random.NextDouble()).ToArray();
var sut = RegressionNeuralNetModel.Load(() => new StringReader(m_regressionNeuralNetModelText));
var predictions = new double[numberOfObservations];
for (int i = 0; i < numberOfObservations; i++)
{
predictions[i] = sut.Predict(observations.Row(i));
}
var evaluator = new MeanSquaredErrorRegressionMetric();
var actual = evaluator.Error(targets, predictions);
Assert.AreEqual(0.0871547675024143, actual, 0.0001);
}
public void ContinuousMunchAugmentor_Augment()
{
var random = new Random(2342);
var data = new F64Matrix(10, 2);
data.Map(() => random.NextDouble());
var sut = new ContinuousMungeAugmentator(0.5, 1.0);
var actual = sut.Agument(data);
var expected = new F64Matrix(new double[] { 0.246329100917247, 0.00372775551105279, 0.797896387892727, 0.441504622549519, 0.585635684703307, 0.227548045212192, 0.254812818139239, 0.387049265851755, 0.722062122878647, 0.888775135804329, 0.714802723710799, 0.900545311114073, 0.643038906922116, 0.907352483788204, 0.316260776164132, 0.658306110025526, 0.343312428492733, 0.0337710422620974, 0.0760259759035082, 0.148426381940221 },
10, 2);
Assert.AreNotEqual(data, actual);
Assert.AreEqual(expected.RowCount, actual.RowCount);
Assert.AreEqual(expected.ColumnCount, actual.ColumnCount);
var expectedData = expected.Data();
var actualData = expected.Data();
for (int i = 0; i < expectedData.Length; i++)
{
Assert.AreEqual(expectedData[i], actualData[i], 0.00001);
}
}
public void NominalMunchAugmentor_Augment()
{
var random = new Random(2342);
var data = new F64Matrix(10, 2);
data.Map(() => random.Next(2));
var sut = new NominalMungeAugmentator(0.5);
var actual = sut.Agument(data);
var expected = new F64Matrix(new double[] { 0, 0, 1, 0, 1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 0, 0 },
10, 2);
Assert.AreNotEqual(data, actual);
Assert.AreEqual(expected.RowCount, actual.RowCount);
Assert.AreEqual(expected.ColumnCount, actual.ColumnCount);
var expectedData = expected.Data();
var actualData = expected.Data();
for (int i = 0; i < expectedData.Length; i++)
{
Assert.AreEqual(expectedData[i], actualData[i], 0.00001);
}
}
public void ClassificationNeuralNetModel_PredictProbability_Single()
{
var numberOfObservations = 500;
var numberOfFeatures = 5;
var numberOfClasses = 5;
var random = new Random(32);
var observations = new F64Matrix(numberOfObservations, numberOfFeatures);
observations.Map(() => random.NextDouble());
var targets = Enumerable.Range(0, numberOfObservations)
.Select(i => (double)random.Next(0, numberOfClasses)).ToArray();
var sut = ClassificationNeuralNetModel.Load(() => new StringReader(m_classificationNeuralNetModelText));
var predictions = new ProbabilityPrediction[numberOfObservations];
for (int i = 0; i < numberOfObservations; i++)
{
predictions[i] = sut.PredictProbability(observations.Row(i));
}
var evaluator = new TotalErrorClassificationMetric <double>();
var actual = evaluator.Error(targets, predictions.Select(p => p.Prediction).ToArray());
Assert.AreEqual(0.762, actual);
}
public NetworkHelper(string inputFileName)
{
var trainingParser = new CsvParser(() => new StreamReader(new FileStream(inputFileName, FileMode.Open)));
var featureNames = trainingParser.EnumerateRows(c => c != TargetName).First().ColumnNameToIndex.Keys
.ToArray();
_trainingObservations = trainingParser.EnumerateRows(featureNames).ToF64Matrix();
_trainingTargets = trainingParser.EnumerateRows(TargetName).ToF64Vector();
_trainingObservations.Map(p => p / 255);
}
public void MatrixAdd_AddInPlace_1()
{
var m = new F64Matrix(4, 3);
var v = new double[] { 1, 1, 1, 1 };
var actual = new F64Matrix(4, 3);
MatrixAdd.AddF64(m, v, actual);
var expected = new F64Matrix(4, 3);
expected.Map(() => 1.0);
Assert.AreEqual(expected, actual);
}
public void MatrixSubtraction_SubtractF64Matrix_1()
{
var m1 = new F64Matrix(4, 3);
var m2 = new F64Matrix(4, 3);
m2.Map(() => 1.0);
var actual = new F64Matrix(4, 3);
MatrixSubtraction.SubtractF64(m1, m2, actual);
var expected = new F64Matrix(4, 3);
expected.Map(() => - 1.0);
Assert.AreEqual(expected, actual);
}
public void BackwardEliminationRegressionEnsembleSelection_Select()
{
var sut = new BackwardEliminationRegressionEnsembleSelection(new MeanSquaredErrorRegressionMetric(),
new MeanRegressionEnsembleStrategy(), 3);
var random = new Random(42);
var observations = new F64Matrix(10, 10);
observations.Map(() => random.Next());
var targets = Enumerable.Range(0, 10).Select(v => random.NextDouble()).ToArray();
var actual = sut.Select(observations, targets);
var expected = new int[] { 2 };
CollectionAssert.AreEqual(expected, actual);
}
public void ForwardSearchRegressionEnsembleSelection_Select_Start_With_2_Models()
{
var sut = new ForwardSearchRegressionEnsembleSelection(new MeanSquaredErrorRegressionMetric(),
new MeanRegressionEnsembleStrategy(), 3, 2, true);
var random = new Random(42);
var observations = new F64Matrix(10, 10);
observations.Map(() => random.Next());
var targets = Enumerable.Range(0, 10).Select(v => random.NextDouble()).ToArray();
var actual = sut.Select(observations, targets);
var expected = new int[3] {
2, 3, 2
};
CollectionAssert.AreEqual(expected, actual);
}
public void ClassificationNeuralNetModel_Predict_Multiple()
{
var numberOfObservations = 500;
var numberOfFeatures = 5;
var numberOfClasses = 5;
var random = new Random(32);
var observations = new F64Matrix(numberOfObservations, numberOfFeatures);
observations.Map(() => random.NextDouble());
var targets = Enumerable.Range(0, numberOfObservations).Select(i => (double)random.Next(0, numberOfClasses)).ToArray();
var sut = ClassificationNeuralNetModel.Load(() => new StringReader(ClassificationNeuralNetModelText));
var predictions = sut.Predict(observations);
var evaluator = new TotalErrorClassificationMetric <double>();
var actual = evaluator.Error(targets, predictions);
Assert.AreEqual(0.762, actual);
}
void CreateData(int numberOfObservations, int numberOfFeatures, Random random, out F64Matrix observations, out double[] targets)
{
observations = new F64Matrix(numberOfObservations, numberOfFeatures);
observations.Map(() => random.NextDouble());
targets = Enumerable.Range(0, numberOfObservations).Select(i => random.NextDouble()).ToArray();
}
