/**
* Copyright (C) 2009 - present by OpenGamma Inc. and the OpenGamma group of companies
*
* Please see distribution for license.
*/
package com.opengamma.analytics.financial.model.volatility.smile.fitting;
import static org.testng.AssertJUnit.assertEquals;
import static org.testng.AssertJUnit.assertTrue;
import java.util.BitSet;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.testng.annotations.Test;
import cern.jet.random.engine.MersenneTwister;
import com.opengamma.analytics.financial.model.option.pricing.analytic.formula.BlackFunctionData;
import com.opengamma.analytics.financial.model.option.pricing.analytic.formula.EuropeanVanillaOption;
import com.opengamma.analytics.financial.model.volatility.smile.function.SABRFormulaData;
import com.opengamma.analytics.financial.model.volatility.smile.function.SABRHaganVolatilityFunction;
import com.opengamma.analytics.math.function.Function1D;
import com.opengamma.analytics.math.statistics.distribution.NormalDistribution;
import com.opengamma.analytics.math.statistics.distribution.ProbabilityDistribution;
import com.opengamma.analytics.math.statistics.leastsquare.LeastSquareResultsWithTransform;
import com.opengamma.util.monitor.OperationTimer;
import com.opengamma.util.test.TestGroup;
/**
* @deprecated This class tests deprecated functionality
*/
@Deprecated
@Test(groups = TestGroup.UNIT)
public class SABRFittingTest {
protected Logger _logger = LoggerFactory.getLogger(SABRFittingTest.class);
protected int _hotspotWarmupCycles = 0;
protected int _benchmarkCycles = 1;
private static final double F = 0.03;
private static final double T = 7.0;
private static final double ALPHA = 0.05;
private static final double BETA = 0.5;
private static double NU = 0.2;
private static double RHO = -0.3;
private static final EuropeanVanillaOption[] OPTIONS;
private static final BlackFunctionData[] DATA;
private static final BlackFunctionData[] NOISY_DATA;
private static double[] STRIKES;
private static double[] ERRORS;
private static final SABRFormulaData SABR_DATA = new SABRFormulaData(ALPHA, BETA, RHO, NU);
private static SABRHaganVolatilityFunction SABR = new SABRHaganVolatilityFunction();
private static ProbabilityDistribution<Double> RANDOM = new NormalDistribution(0, 1, new MersenneTwister(12));
private static final SABRNonLinearLeastSquareFitter NLSS = new SABRNonLinearLeastSquareFitter(SABR);
// private static final SABRConjugateGradientLeastSquareFitter CG = new SABRConjugateGradientLeastSquareFitter(SABR);
static {
STRIKES = new double[] {0.005, 0.01, 0.02, 0.03, 0.04, 0.05, 0.07 };
final int n = STRIKES.length;
DATA = new BlackFunctionData[n];
NOISY_DATA = new BlackFunctionData[n];
ERRORS = new double[n];
OPTIONS = new EuropeanVanillaOption[n];
for (int i = 0; i < n; i++) {
final EuropeanVanillaOption option = new EuropeanVanillaOption(STRIKES[i], T, true);
OPTIONS[i] = option;
final double vol = SABR.getVolatilityFunction(option, F).evaluate(SABR_DATA);
DATA[i] = new BlackFunctionData(F, 1, vol);
ERRORS[i] = 0.01;
NOISY_DATA[i] = new BlackFunctionData(F, 1, vol + ERRORS[i] * RANDOM.nextRandom());
}
}
@Test
public void testExactFitNLSS() {
final BitSet fixed = new BitSet();
final double[] start = new double[] {0.03, 0.4, 0.1, 0.2 };
final LeastSquareResultsWithTransform results = NLSS.getFitResult(OPTIONS, DATA, ERRORS, start, fixed);
final double[] res = results.getModelParameters().getData();
final double eps = 1e-4;
assertEquals(ALPHA, res[0], eps);
assertEquals(BETA, res[1], eps);
assertEquals(RHO, res[2], eps);
assertEquals(NU, res[3], eps);
assertEquals(0.0, results.getChiSq(), eps);
}
//FIXME: test doesn't pass at all
// @SuppressWarnings("unused")
// @Test
// public void testExactFitCG() {
// final BitSet fixed = new BitSet();
// final double[] start = new double[] {0.03, 0.4, 0.1, 0.2};
// final LeastSquareResults results = CG.getFitResult(OPTIONS, DATA, ERRORS, start, fixed);
// final double[] res = results.getParameters().getData();
// final double eps = 1e-4;
// // assertEquals(ALPHA, res[0], eps);
// // assertEquals(BETA, res[1], eps);
// // assertEquals(NU, res[2], eps);
// // assertEquals(RHO, res[3], eps);
// // assertEquals(0.0, results.getChiSq(), eps);
// }
@Test
public void testSmileGenerationTime() {
final SABRHaganVolatilityFunction sabr = new SABRHaganVolatilityFunction();
final SABRFormulaData data = new SABRFormulaData(ALPHA, BETA, RHO, NU);
final EuropeanVanillaOption option = new EuropeanVanillaOption(0.9 * F, T, true);
final Function1D<SABRFormulaData, Double> f = sabr.getVolatilityFunction(option, F);
double x = 0;
for (int i = 0; i < _hotspotWarmupCycles; i++) {
x += f.evaluate(data);
}
if (_benchmarkCycles > 0) {
final OperationTimer timer = new OperationTimer(_logger, "processing {} cycles generating SABR smile", _benchmarkCycles);
for (int i = 0; i < _benchmarkCycles; i++) {
x += f.evaluate(data);
}
timer.finished();
}
}
@Test
public void testFitTime() {
assertFitTime(NLSS, "non-linear least squares");
// testFitTime(CG, "conjugate gradient");
}
private void assertFitTime(final LeastSquareSmileFitter fitter, final String name) {
final BitSet fixed = new BitSet();
final double[] start = new double[] {0.03, 0.4, 0.1, 0.2 };
for (int i = 0; i < _hotspotWarmupCycles; i++) {
fitter.getFitResult(OPTIONS, DATA, ERRORS, start, fixed);
}
if (_benchmarkCycles > 0) {
final OperationTimer timer = new OperationTimer(_logger, "processing {} cycles fitting with {}", _benchmarkCycles, name);
for (int i = 0; i < _benchmarkCycles; i++) {
fitter.getFitResult(OPTIONS, DATA, ERRORS, start, fixed);
}
timer.finished();
}
}
@Test
public void testNoisyFitNLSS() {
final BitSet fixed = new BitSet();
fixed.set(1, true);
final double[] start = new double[] {0.03, 0.5, 0.1, 0.2 };
final LeastSquareResultsWithTransform results = NLSS.getFitResult(OPTIONS, NOISY_DATA, ERRORS, start, fixed);
assertTrue(results.getChiSq() < 10.0);
final double[] res = results.getModelParameters().getData();
assertEquals(ALPHA, res[0], 1e-3);
assertEquals(BETA, res[1], 1e-7);
assertEquals(RHO, res[2], 1e-1);
assertEquals(NU, res[3], 1e-2);
}
//FIXME: tests don't pass at all
// @SuppressWarnings("unused")
// @Test
// public void testNoisyFitCG() {
// final BitSet fixed = new BitSet();
// fixed.set(1, true);
// final double[] start = new double[] {0.03, 0.5, 0.1, 0.2};
// final LeastSquareResults results = CG.getFitResult(OPTIONS, NOISY_DATA, ERRORS, start, fixed);
// // assertTrue(results.getChiSq() < 10.0);
// // final double[] res = results.getParameters().getData();
// // assertEquals(ALPHA, res[0], 1e-3);
// // assertEquals(BETA, res[1], 1e-7);
// // assertEquals(RHO, res[3], 1e-1);
// // assertEquals(NU, res[2], 1e-2);
// }
}