public DVPLI.EstimationResult Estimate(List <object> data, DVPLI.IEstimationSettings settings = null, DVPLI.IController controller = null, Dictionary <string, object> properties = null)
        {
            DVPLI.InterestRateMarketData irmd = data[0] as DVPLI.InterestRateMarketData;

            //Date today = new Date(15, Month.February, 2002);
            //Date settlement = new Date(19, Month.February, 2002);
            Settings.setEvaluationDate(irmd.Date);



            Handle <YieldTermStructure> termStructure = new Handle <YieldTermStructure>(new Utilities.ZeroRateFunction(irmd.Date, irmd.ZRMarketDates, irmd.ZRMarket));

            //termStructure.link
            HullWhite model = new HullWhite(termStructure);


            IborIndex index = new Euribor6M(termStructure);

            IPricingEngine engine = new JamshidianSwaptionEngine(model);

            List <CalibrationHelper> swaptions = new List <CalibrationHelper>();

            for (int i = 0; i < irmd.SwapDates.Length; i++)
            {
                for (int j = 0; j < irmd.SwapDuration.Length; j++)
                {
                    Quote             vol    = new SimpleQuote(irmd.SwaptionsVolatility[j, i]);
                    CalibrationHelper helper =
                        new SwaptionHelper(new Period((int)irmd.SwapDates[i], TimeUnit.Years),
                                           new Period((int)irmd.SwapDuration[j], TimeUnit.Years),
                                           new Handle <Quote>(vol),
                                           index,
                                           new Period(1, TimeUnit.Years),
                                           new Thirty360(),
                                           new Actual360(),
                                           termStructure,
                                           false);
                    helper.setPricingEngine(engine);
                    swaptions.Add(helper);
                }
            }

            // Set up the optimization problem
            LevenbergMarquardt optimizationMethod = new LevenbergMarquardt(1.0e-8, 1.0e-8, 1.0e-8);
            EndCriteria        endCriteria        = new EndCriteria(10000, 100, 1e-6, 1e-8, 1e-8);

            //Optimize
            model.calibrate(swaptions, optimizationMethod, endCriteria, new Constraint(), new List <double>());
            EndCriteria.Type ecType = model.endCriteria();


            Vector xMinCalculated = model.parameters();
            double yMinCalculated = model.value(xMinCalculated, swaptions);
            Vector xMinExpected   = new Vector(2);

            double yMinExpected = model.value(xMinExpected, swaptions);

            DVPLI.EstimationResult r = new DVPLI.EstimationResult(new string[] { "Alpha", "Sigma" }, new double[] { xMinCalculated[0], xMinCalculated[1] });
            return(r);
        }
        public void testCachedHullWhite()
        {
            //("Testing Hull-White calibration against cached values...");

            Date today      = new Date(15, Month.February, 2002);
            Date settlement = new Date(19, Month.February, 2002);

            Settings.setEvaluationDate(today);
            Handle <YieldTermStructure> termStructure =
                new Handle <YieldTermStructure>(Utilities.flatRate(settlement, 0.04875825, new Actual365Fixed()));
            //termStructure.link
            HullWhite model = new HullWhite(termStructure);

            CalibrationData[] data = { new CalibrationData(1, 5, 0.1148),
                                       new CalibrationData(2, 4, 0.1108),
                                       new CalibrationData(3, 3, 0.1070),
                                       new CalibrationData(4, 2, 0.1021),
                                       new CalibrationData(5, 1, 0.1000) };
            IborIndex         index = new Euribor6M(termStructure);

            IPricingEngine engine = new JamshidianSwaptionEngine(model);

            List <CalibrationHelper> swaptions = new List <CalibrationHelper>();

            for (int i = 0; i < data.Length; i++)
            {
                Quote             vol    = new SimpleQuote(data[i].volatility);
                CalibrationHelper helper =
                    new SwaptionHelper(new Period(data[i].start, TimeUnit.Years),
                                       new Period(data[i].length, TimeUnit.Years),
                                       new Handle <Quote>(vol),
                                       index,
                                       new Period(1, TimeUnit.Years),
                                       new Thirty360(),
                                       new Actual360(),
                                       termStructure,
                                       false);
                helper.setPricingEngine(engine);
                swaptions.Add(helper);
            }

            // Set up the optimization problem
            // Real simplexLambda = 0.1;
            // Simplex optimizationMethod(simplexLambda);
            LevenbergMarquardt optimizationMethod = new LevenbergMarquardt(1.0e-8, 1.0e-8, 1.0e-8);
            EndCriteria        endCriteria        = new EndCriteria(10000, 100, 1e-6, 1e-8, 1e-8);

            //Optimize
            model.calibrate(swaptions, optimizationMethod, endCriteria, new Constraint(), new List <double>());
            EndCriteria.Type ecType = model.endCriteria();

            // Check and print out results
            #if QL_USE_INDEXED_COUPON
            double cachedA = 0.0488199, cachedSigma = 0.00593579;
            #else
            double cachedA = 0.0488565, cachedSigma = 0.00593662;
            #endif
            double tolerance = 1.120e-5;
            //double tolerance = 1.0e-6;
            Vector xMinCalculated = model.parameters();
            double yMinCalculated = model.value(xMinCalculated, swaptions);
            Vector xMinExpected   = new Vector(2);
            xMinExpected[0] = cachedA;
            xMinExpected[1] = cachedSigma;
            double yMinExpected = model.value(xMinExpected, swaptions);
            if (Math.Abs(xMinCalculated[0] - cachedA) > tolerance ||
                Math.Abs(xMinCalculated[1] - cachedSigma) > tolerance)
            {
                Assert.Fail("Failed to reproduce cached calibration results:\n"
                            + "calculated: a = " + xMinCalculated[0] + ", "
                            + "sigma = " + xMinCalculated[1] + ", "
                            + "f(a) = " + yMinCalculated + ",\n"
                            + "expected:   a = " + xMinExpected[0] + ", "
                            + "sigma = " + xMinExpected[1] + ", "
                            + "f(a) = " + yMinExpected + ",\n"
                            + "difference: a = " + (xMinCalculated[0] - xMinExpected[0]) + ", "
                            + "sigma = " + (xMinCalculated[1] - xMinExpected[1]) + ", "
                            + "f(a) = " + (yMinCalculated - yMinExpected) + ",\n"
                            + "end criteria = " + ecType);
            }
        }