public void SplineSlopeConditionsTest()
{
ClString name = "SplineSlopeConditionsTest";
BeginTest(name);
// Testing data: CallGeomBrownian; Linear, Parabolic, Exponential Data; PayoffButterfly, PayoffCallOption.
ApproximationTestsHelper test = new ApproximationTestsHelper(Log, Context, name, ConsoleOutput);
foreach (ClString testName in test.GetInputFileNames(GetType().FullName, "Slope"))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
// Read slope.
double slopeLeft = 0;
double slopeRight = 0;
var n = reference.Length;
try
{
fileName = testName + "Slope.csv";
var matrix = new ClCsvMatrix(Context.AsFileContext.LoadText(fileName));
slopeLeft = matrix[0, 0].ToDouble();
slopeRight = matrix[0, 1].ToDouble();
}
catch (Exception)
{
// if slopes are not set, calculates slopes by first and last two points of reference points.
slopeLeft = (reference[1].Value - reference[0].Value) / (reference[1].X[0] - reference[0].X[0]);
slopeRight = (reference[n - 1].Value - reference[n - 2].Value) / (reference[n - 1].X[0] - reference[n - 2].X[0]);
}
// Initialise spline and calculate spline coefficients.
var spline = new ClSplineCubicSmoothSlope1D(new Spline1DBuilder())
{
Params =
{
SlopeLeft = slopeLeft,
SlopeRight = slopeRight
}
};
spline.LoadData(data);
test.CalculationTime.Restart();
spline.Calculate();
test.CalculationTime.Stop();
// Store results and complete test.
test.Print(String.Format("Optimal lambda: {0}", spline.Params.SmoothParam));
if (PlotResults)
{
test.StoreSpline1D(spline, reference);
}
test.Print("Completed " + name + " for " + testName);
}
test.FinishTest(PrintPerformance);
}
public void SplineEndValueTest()
{
ClString name = "SplineEndValueTest";
BeginTest(name);
// Testing data: Call, Put, Risk Reversal, Straddle, Strangle, Bull, Bear, Butterfly and Barrier (Down-and-in call, Down-and-out call, Up-and-in call, Up-and-out call ) Options.
ApproximationTestsHelper test = new ApproximationTestsHelper(Log, Context, name, ConsoleOutput);
foreach (ClString testName in test.GetInputFileNames(GetType().FullName, "Value"))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
// Read end values.
double valueLeft = 0;
double valueRight = 0;
var n = reference.Length;
try
{
fileName = testName + "Value.csv";
var matrix = new ClCsvMatrix(Context.AsFileContext.LoadText(fileName));
valueLeft = matrix[0, 0].ToDouble();
valueRight = matrix[0, 1].ToDouble();
}
catch (Exception)
{
// set reference values at end points if end values are not set.
valueLeft = reference[0].Value;
valueRight = reference[n - 1].Value;
}
// Initialise spline and calculate spline coefficients.
Spline1DBuilder builder = new Spline1DBuilder();
builder.ValueLeft = valueLeft;
builder.ValueRight = valueRight;
var spline = new ClSplineCubicSmoothEndValue1D(builder);
spline.LoadData(data);
test.CalculationTime.Restart();
spline.Calculate();
test.CalculationTime.Stop();
// Store results and complete test.
test.Print(String.Format("Optimal lambda: {0}", spline.Params.SmoothParam));
if (PlotResults)
{
test.StoreSpline1D(spline, reference);
}
test.Print("Completed " + name + " for " + testName);
}
test.FinishTest(PrintPerformance);
}
public void ConvexSplineTest()
{
ClString name = "ConvexSplineTest";
BeginTest(name);
const double convexityConstraint = 0.5;
const double step = 0.1;
var test = new ConvexSplineTest(Log, Context, name, ConsoleOutput);
// Testing data: CallGeomBrownian; FunctionExp(X), FunctionXPow2, FunctionXPow4; PayoffCallOption.
foreach (ClString testName in test.GetInputFileNames(GetType().FullName))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
// Initialise spline and calculate spline coefficients.
test.Print(String.Format("Calculating spline."));
var convexSpline = new ClConvexSpline1D
{
ConvexityConstraint = convexityConstraint,
Step = step
};
convexSpline.LoadData(data);
convexSpline.Extrapolate = false;
test.CalculationTime.Restart();
convexSpline.Calculate();
test.CalculationTime.Stop();
var dat = convexSpline.GridData(step);
var smoothingSpline = new ClSplineCubicSmooth1D(new Spline1DBuilder());
smoothingSpline.LoadData(data);
smoothingSpline.Calculate();
if (PlotResults)
{
test.StoreSpline1D(smoothingSpline, convexSpline, data, reference);
}
Log.Result(String.Format(name + " for " + testName + " completed." + "{0}" +
"Number of points: " + data.Length + ";{0}" +
"Processing time: " + test.CalculationTime.ElapsedMilliseconds + ";{0}" +
"Deviation between spline and reference: " + ClApproximUtils.DeviationBetween(convexSpline, reference) + ";{0}" +
"Convex measure: " + convexSpline.ConvexMeasure(dat) + ";{0}", Environment.NewLine));
test.CalculationTime.Reset();
}
Log.Result(name + " completed");
test.FinishTest(PrintPerformance);
}
public void MonotoneSplineTest()
{
ClString name = "MonotoneSplineTest";
BeginTest(name);
var test = new MonotoneSplineTest(Log, Context, name, ConsoleOutput);
// Testing data: CallGeomBrownian; f(x) = x, F(x) = x^4 + 2*x^3 + 5*x*x - 6*x + 7,
// f(x) = sin(x), f(x) = 1/2 (x-cos(x) sin(x)), f(x) = x+e^-x, etc.
foreach (ClString testName in test.GetInputFileNames(GetType().FullName))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
// Initialise spline and calculate spline coefficients.
test.Print(String.Format("Calculating spline."));
var monotoneSpline = new ClMonotoneSpline1D();
test.CalculationTime.Restart();
monotoneSpline.LoadData(data);
monotoneSpline.Calculate();
test.CalculationTime.Stop();
var gridData = monotoneSpline.ValueOnGrid(0.01);
var dat = new ClPoint[gridData.RowCount];
for (var i = 0; i < dat.Length; i++)
{
dat[i] = new ClPoint(gridData[i, 0], gridData[i, 1]);
}
var smoothingSpline = new ClSplineCubicSmooth1D(new Spline1DBuilder());
smoothingSpline.LoadData(data);
smoothingSpline.Calculate();
if (PlotResults)
{
test.StoreSpline1D(smoothingSpline, monotoneSpline, data, reference);
}
Log.Result(String.Format(name + " for " + testName + " completed." + "{0}" +
"Number of points: " + data.Length + ";{0}" +
"Processing time: " + test.CalculationTime.ElapsedMilliseconds + ";{0}" +
"Deviation between spline and reference: " + ClApproximUtils.DeviationBetween(monotoneSpline, reference) + ";{0}", Environment.NewLine));
}
Log.Result(name + " completed");
test.FinishTest(PrintPerformance);
}
public void SplineOptimalSmoothTest()
{
ClString name = "SplineOptimalSmoothTest";
BeginTest(name);
// Testing data: CallGeomBrownian; Linear, Parabolic, Exponential Data; PayoffButterfly, PayoffCallOption.
var test = new ApproximationTestsHelper(Log, Context, name, ConsoleOutput);
foreach (ClString testName in test.GetInputFileNames(GetType().FullName))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
// Initialise spline and calculate spline coefficients.
var spline = new ClSplineCubicOptimalSmooth1D(new Spline1DBuilder())
{
OptimSteer =
{
StartSeed = StartSeed,
Method = ClSplineCubicOptimalSmooth1D.OptimizationSteering.UseMethod.CvGoldenSection
}
};
// TODO Switch to CvGoldenSection, CvScan, NcpScan or NcpGoldenSection optimization.
spline.LoadData(data);
test.CalculationTime.Restart();
spline.Calculate();
test.CalculationTime.Stop();
// Store results and complete test.
test.Print(String.Format("Optimal lambda: {0}", spline.Params.SmoothParam));
if (PlotResults)
{
test.StoreSpline1D(spline, reference);
}
test.Print("Completed " + name + " for " + testName);
}
test.FinishTest(PrintPerformance);
}
public void KernelRegressionTest()
{
ClString name = "KernelRegressionTest";
BeginTest(name);
// Testing data: CallGeomBrownian; Linear, Parabolic, Exponential Data; PayoffButterfly, PayoffCallOption.
NadarayaWatsonKernelRegressionTest test = new NadarayaWatsonKernelRegressionTest(Log, Context, name, ConsoleOutput);
foreach (ClString testName in test.GetInputFileNames(GetType().FullName))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
test.TotalPointsNumber += data.Length;
// Initialise spline and calculate spline coefficients.
test.CalculationTime.Restart();
ClNadarayaWatsonKernelRegression spline = new ClNadarayaWatsonKernelRegression(new ClKernelRegressionParams());
//spline.OptimSteer.StartSeed = startSeed;
//// TODO Switch to CvGoldenSection, CvScan, NcpScan or NcpGoldenSection optimization.
//spline.OptimSteer.Method = ClSplineCubicOptimalSmooth1D.OptimizationSteering.UseMethod.CvGoldenSection;
spline.LoadData(data);
//spline.Calculate();
test.CalculationTime.Stop();
Array.Sort(data, new ClPointComparerX());
ClCsvMatrix gridData = test.ValueOnGrid(spline, data[0].X[0], data[data.Length - 1].X[0], 0.01);
// Store results and complete test.
if (PlotResults)
{
test.StoreNadarayaWatson1D(spline, data, reference, gridData);
}
test.Print("Completed " + name + " for " + testName);
}
test.FinishTest(PrintPerformance);
}
public void SplineTest()
{
ClString name = "SplineTest";
BeginTest(name);
// Testing data: Call, Put, Risk Reversal, Straddle, Strangle, Bull, Bear, Butterfly and Barrier (Down-and-in call, Down-and-out call, Up-and-in call, Up-and-out call ) Options.
ApproximationTestsHelper test = new ApproximationTestsHelper(Log, Context, name, ConsoleOutput);
foreach (ClString testName in test.GetInputFileNames(GetType().FullName))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
ClSplineCubicSmooth1D spline = new ClSplineCubicSmooth1D(new Spline1DBuilder());
spline.LoadData(data);
test.CalculationTime.Restart();
spline.Calculate();
test.CalculationTime.Stop();
// Store results and complete test.
test.Print(String.Format("Optimal lambda: {0}", spline.Params.SmoothParam));
if (PlotResults)
{
test.StoreSpline1D(spline, reference);
}
test.Print("Completed " + name + " for " + testName);
}
test.FinishTest(PrintPerformance);
}
public void AMCRegressionTest()
{
ClString name = "AMCRegressionTest";
BeginTest(name);
// Testing data: Call, Put, Risk Reversal, Straddle, Strangle, Bull, Bear, Butterfly and Barrier (Down-and-in call, Down-and-out call, Up-and-in call, Up-and-out call ) Options.
ApproximationTestsHelper test = new ApproximationTestsHelper(Log, Context, name, ConsoleOutput);
foreach (ClString testName in test.GetInputFileNames(GetType().FullName))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
// initial data for polynomial regression
double[] dataX = new double[data.Length];
for (int pos = 0; pos < data.Length; pos++)
{
dataX[pos] = data[pos].X[0];
}
ClDoubleArray explanatoryVariables = new ClDoubleArray(dataX);
double[] dataV = new double[data.Length];
for (int pos = 0; pos < data.Length; pos++)
{
dataV[pos] = data[pos].Value;
}
ClDoubleArray exposureDiscounted = new ClDoubleArray(dataV);
// Initialise polynomial regression and calculate it
var amcBuilder = new ClPolynomialLsqAmcBuilder
{
Sdf = new ClDoubleArray(data.Length, 1.0),
ExplanatoryVariables = explanatoryVariables,
PolynomialPower = 3
};
IClAmc amc = amcBuilder.Build <IClAmc>();
test.CalculationTime.Restart();
ClDoubleArray result = new ClDoubleArray(amc.Evaluate(exposureDiscounted));
test.CalculationTime.Stop();
ClPoint[] valuesAt = new ClPoint[data.Length];
for (int pos = 0; pos < data.Length; pos++)
{
valuesAt[pos] = new ClPoint(data[pos].X, result[pos]);
}
// Initialise spline and calculate spline coefficients.
var spline = new ClSplineCubicSmooth1D(new Spline1DBuilder());
spline.LoadData(data);
spline.Calculate();
// Store results and complete test.
if (PlotResults)
{
test.StoreSpline1D(spline, valuesAt, data, reference);
}
test.Print("Completed " + name + " for " + testName);
}
test.FinishTest(PrintPerformance);
}
public void SplineBoundTest()
{
ClString name = "SplineBoundTest";
BeginTest(name);
// Testing data: CallGeomBrownian, PayoffButterfly, PayoffCallOption.
var test = new ApproximationTestsHelper(Log, Context, name, ConsoleOutput);
foreach (ClString testName in test.GetInputFileNames(GetType().FullName, "Bounds"))
{
test.StartTest(testName);
// Read input data.
ClString fileName = testName + ".csv";
ClPoint[] data, reference;
test.ReadCSVInputData(fileName, out data, out reference);
// Read bounds.
ClDouble boundMin;
ClDouble boundMax;
const bool fromFile = false;
// TODO Switch to true to load bounds from file
if (fromFile)
{
fileName = testName + "Bounds.csv";
ClCsvMatrix matrix = new ClCsvMatrix(Context.AsFileContext.LoadText(fileName));
boundMin = matrix[0, 0].ToDouble();
boundMax = matrix[0, 1].ToDouble();
}
else
{
var refer = new double[reference.Length];
for (var i = 0; i < refer.Length; i++)
{
refer[i] = reference[i].Value;
}
boundMin = refer.Min();
boundMax = refer.Max();
}
// Initialise spline and calculate spline coefficients.
var splineParams = new Spline1DBuilder {
Min = boundMin, Max = boundMax
};
var spline = new ClSplineCubicSmooth1D(splineParams);
spline.LoadData(data);
test.CalculationTime.Restart();
spline.Calculate();
test.CalculationTime.Stop();
// Store results and complete test.
if (PlotResults)
{
test.StoreSpline1D(spline, reference);
}
test.Print("Completed " + name + " for " + testName);
}
test.FinishTest(PrintPerformance);
}
