protected void CalculateTimeSeriesResults()
{
OnlineCalculatorError errorState;
double trainingMean = ProblemData.TrainingIndices.Any() ? ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TrainingIndices).Average() : double.NaN;
var meanModel = new ConstantModel(trainingMean, ProblemData.TargetVariable);
double alpha, beta;
IEnumerable <double> trainingStartValues = ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TrainingIndices.Select(r => r - 1).Where(r => r > 0)).ToList();
OnlineLinearScalingParameterCalculator.Calculate(ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TrainingIndices.Where(x => x > 0)), trainingStartValues, out alpha, out beta, out errorState);
var AR1model = new TimeSeriesPrognosisAutoRegressiveModel(ProblemData.TargetVariable, new double[] { beta }, alpha);
#region Calculate training quality measures
if (ProblemData.TrainingIndices.Any())
{
IEnumerable <double> trainingTargetValues = ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TrainingIndices).ToList();
IEnumerable <double> trainingEstimatedValues = EstimatedTrainingValues.ToList();
IEnumerable <double> trainingMeanModelPredictions = meanModel.GetEstimatedValues(ProblemData.Dataset, ProblemData.TrainingIndices).ToList();
IEnumerable <double> trainingAR1ModelPredictions = AR1model.GetEstimatedValues(ProblemData.Dataset, ProblemData.TrainingIndices).ToList();
TrainingDirectionalSymmetry = OnlineDirectionalSymmetryCalculator.Calculate(trainingTargetValues.First(), trainingTargetValues, trainingEstimatedValues, out errorState);
TrainingDirectionalSymmetry = errorState == OnlineCalculatorError.None ? TrainingDirectionalSymmetry : 0.0;
TrainingWeightedDirectionalSymmetry = OnlineWeightedDirectionalSymmetryCalculator.Calculate(trainingTargetValues.First(), trainingTargetValues, trainingEstimatedValues, out errorState);
TrainingWeightedDirectionalSymmetry = errorState == OnlineCalculatorError.None ? TrainingWeightedDirectionalSymmetry : 0.0;
TrainingTheilsUStatisticAR1 = OnlineTheilsUStatisticCalculator.Calculate(trainingTargetValues.First(), trainingTargetValues, trainingAR1ModelPredictions, trainingEstimatedValues, out errorState);
TrainingTheilsUStatisticAR1 = errorState == OnlineCalculatorError.None ? TrainingTheilsUStatisticAR1 : double.PositiveInfinity;
TrainingTheilsUStatisticMean = OnlineTheilsUStatisticCalculator.Calculate(trainingTargetValues.First(), trainingTargetValues, trainingMeanModelPredictions, trainingEstimatedValues, out errorState);
TrainingTheilsUStatisticMean = errorState == OnlineCalculatorError.None ? TrainingTheilsUStatisticMean : double.PositiveInfinity;
}
#endregion
#region Calculate test quality measures
if (ProblemData.TestIndices.Any())
{
IEnumerable <double> testTargetValues = ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TestIndices).ToList();
IEnumerable <double> testEstimatedValues = EstimatedTestValues.ToList();
IEnumerable <double> testMeanModelPredictions = meanModel.GetEstimatedValues(ProblemData.Dataset, ProblemData.TestIndices).ToList();
IEnumerable <double> testAR1ModelPredictions = AR1model.GetEstimatedValues(ProblemData.Dataset, ProblemData.TestIndices).ToList();
TestDirectionalSymmetry = OnlineDirectionalSymmetryCalculator.Calculate(testTargetValues.First(), testTargetValues, testEstimatedValues, out errorState);
TestDirectionalSymmetry = errorState == OnlineCalculatorError.None ? TestDirectionalSymmetry : 0.0;
TestWeightedDirectionalSymmetry = OnlineWeightedDirectionalSymmetryCalculator.Calculate(testTargetValues.First(), testTargetValues, testEstimatedValues, out errorState);
TestWeightedDirectionalSymmetry = errorState == OnlineCalculatorError.None ? TestWeightedDirectionalSymmetry : 0.0;
TestTheilsUStatisticAR1 = OnlineTheilsUStatisticCalculator.Calculate(testTargetValues.First(), testTargetValues, testAR1ModelPredictions, testEstimatedValues, out errorState);
TestTheilsUStatisticAR1 = errorState == OnlineCalculatorError.None ? TestTheilsUStatisticAR1 : double.PositiveInfinity;
TestTheilsUStatisticMean = OnlineTheilsUStatisticCalculator.Calculate(testTargetValues.First(), testTargetValues, testMeanModelPredictions, testEstimatedValues, out errorState);
TestTheilsUStatisticMean = errorState == OnlineCalculatorError.None ? TestTheilsUStatisticMean : double.PositiveInfinity;
}
#endregion
}
protected OnlineTheilsUStatisticCalculator(OnlineTheilsUStatisticCalculator original, Cloner cloner)
: base(original, cloner)
{
squaredErrorMeanCalculator = cloner.Clone(original.squaredErrorMeanCalculator);
unbiasedEstimatorMeanCalculator = cloner.Clone(original.unbiasedEstimatorMeanCalculator);
}
private void CalculateTestPrognosisResults()
{
OnlineCalculatorError errorState;
var problemData = Solution.ProblemData;
if (!problemData.TestIndices.Any())
{
return;
}
var model = Solution.Model;
var testHorizions = problemData.TestIndices.Select(r => Math.Min(testHorizon, problemData.TestPartition.End - r)).ToList();
IEnumerable <IEnumerable <double> > testTargetValues = problemData.TestIndices.Zip(testHorizions, Enumerable.Range).Select(r => problemData.Dataset.GetDoubleValues(problemData.TargetVariable, r)).ToList();
IEnumerable <IEnumerable <double> > testEstimatedValues = model.GetPrognosedValues(problemData.Dataset, problemData.TestIndices, testHorizions).ToList();
IEnumerable <double> testStartValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TestIndices.Select(r => r - 1).Where(r => r > 0)).ToList();
IEnumerable <double> originalTestValues = testTargetValues.SelectMany(x => x).ToList();
IEnumerable <double> estimatedTestValues = testEstimatedValues.SelectMany(x => x).ToList();
double testMSE = OnlineMeanSquaredErrorCalculator.Calculate(originalTestValues, estimatedTestValues, out errorState);
PrognosisTestMeanSquaredError = errorState == OnlineCalculatorError.None ? testMSE : double.NaN;
double testMAE = OnlineMeanAbsoluteErrorCalculator.Calculate(originalTestValues, estimatedTestValues, out errorState);
PrognosisTestMeanAbsoluteError = errorState == OnlineCalculatorError.None ? testMAE : double.NaN;
double testR = OnlinePearsonsRCalculator.Calculate(originalTestValues, estimatedTestValues, out errorState);
PrognosisTestRSquared = errorState == OnlineCalculatorError.None ? testR * testR : double.NaN;
double testRelError = OnlineMeanAbsolutePercentageErrorCalculator.Calculate(originalTestValues, estimatedTestValues, out errorState);
PrognosisTestRelativeError = errorState == OnlineCalculatorError.None ? testRelError : double.NaN;
double testNMSE = OnlineNormalizedMeanSquaredErrorCalculator.Calculate(originalTestValues, estimatedTestValues, out errorState);
PrognosisTestNormalizedMeanSquaredError = errorState == OnlineCalculatorError.None ? testNMSE : double.NaN;
double testME = OnlineMeanErrorCalculator.Calculate(originalTestValues, estimatedTestValues, out errorState);
PrognosisTestMeanError = errorState == OnlineCalculatorError.None ? testME : double.NaN;
PrognosisTestDirectionalSymmetry = OnlineDirectionalSymmetryCalculator.Calculate(testStartValues, testTargetValues, testEstimatedValues, out errorState);
PrognosisTestDirectionalSymmetry = errorState == OnlineCalculatorError.None ? PrognosisTestDirectionalSymmetry : 0.0;
PrognosisTestWeightedDirectionalSymmetry = OnlineWeightedDirectionalSymmetryCalculator.Calculate(testStartValues, testTargetValues, testEstimatedValues, out errorState);
PrognosisTestWeightedDirectionalSymmetry = errorState == OnlineCalculatorError.None ? PrognosisTestWeightedDirectionalSymmetry : 0.0;
if (problemData.TrainingIndices.Any())
{
//mean model
double trainingMean = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices).Average();
var meanModel = new ConstantModel(trainingMean);
//AR1 model
double alpha, beta;
IEnumerable <double> trainingStartValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices.Select(r => r - 1).Where(r => r > 0)).ToList();
OnlineLinearScalingParameterCalculator.Calculate(problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices.Where(x => x > 0)), trainingStartValues, out alpha, out beta, out errorState);
var AR1model = new TimeSeriesPrognosisAutoRegressiveModel(problemData.TargetVariable, new double[] { beta }, alpha);
IEnumerable <IEnumerable <double> > testMeanModelPredictions = meanModel.GetPrognosedValues(problemData.Dataset, problemData.TestIndices, testHorizions).ToList();
IEnumerable <IEnumerable <double> > testAR1ModelPredictions = AR1model.GetPrognosedValues(problemData.Dataset, problemData.TestIndices, testHorizions).ToList();
PrognosisTestTheilsUStatisticAR1 = OnlineTheilsUStatisticCalculator.Calculate(testStartValues, testTargetValues, testAR1ModelPredictions, testEstimatedValues, out errorState);
PrognosisTestTheilsUStatisticAR1 = errorState == OnlineCalculatorError.None ? PrognosisTestTheilsUStatisticAR1 : double.PositiveInfinity;
PrognosisTestTheilsUStatisticMean = OnlineTheilsUStatisticCalculator.Calculate(testStartValues, testTargetValues, testMeanModelPredictions, testEstimatedValues, out errorState);
PrognosisTestTheilsUStatisticMean = errorState == OnlineCalculatorError.None ? PrognosisTestTheilsUStatisticMean : double.PositiveInfinity;
}
}
public static double Calculate(double startValue, IEnumerable<double> actualContinuation, IEnumerable<double> referenceContinuation, IEnumerable<double> predictedContinuation, out OnlineCalculatorError errorState) {
OnlineTheilsUStatisticCalculator calculator = new OnlineTheilsUStatisticCalculator();
calculator.Add(startValue, actualContinuation, referenceContinuation, predictedContinuation);
errorState = calculator.ErrorState;
return calculator.TheilsUStatistic;
}
public static double Calculate(IEnumerable<double> startValues, IEnumerable<IEnumerable<double>> actualContinuations, IEnumerable<IEnumerable<double>> referenceContinuations, IEnumerable<IEnumerable<double>> predictedContinuations, out OnlineCalculatorError errorState) {
IEnumerator<double> startValueEnumerator = startValues.GetEnumerator();
IEnumerator<IEnumerable<double>> actualContinuationsEnumerator = actualContinuations.GetEnumerator();
IEnumerator<IEnumerable<double>> referenceContinuationsEnumerator = referenceContinuations.GetEnumerator();
IEnumerator<IEnumerable<double>> predictedContinuationsEnumerator = predictedContinuations.GetEnumerator();
OnlineTheilsUStatisticCalculator calculator = new OnlineTheilsUStatisticCalculator();
// always move forward all enumerators (do not use short-circuit evaluation!)
while (startValueEnumerator.MoveNext() & actualContinuationsEnumerator.MoveNext() & referenceContinuationsEnumerator.MoveNext() & predictedContinuationsEnumerator.MoveNext()) {
calculator.Add(startValueEnumerator.Current, actualContinuationsEnumerator.Current, referenceContinuationsEnumerator.Current, predictedContinuationsEnumerator.Current);
if (calculator.ErrorState != OnlineCalculatorError.None) break;
}
// check if all enumerators are at the end to make sure both enumerations have the same length
if (calculator.ErrorState == OnlineCalculatorError.None &&
(startValueEnumerator.MoveNext() || actualContinuationsEnumerator.MoveNext() || referenceContinuationsEnumerator.MoveNext() || predictedContinuationsEnumerator.MoveNext())) {
throw new ArgumentException("Number of elements in startValues, actualContinuations, referenceContinuation and estimatedValues predictedContinuations doesn't match.");
} else {
errorState = calculator.ErrorState;
return calculator.TheilsUStatistic;
}
}