public JarrowRudd(StochasticProcess1D process, double end, int steps, double strike)
: base(process, end, steps)
{
// drift removed
up_ = process.stdDeviation(0.0, x0_, dt_);
}
public TrinomialTree(StochasticProcess1D process,
TimeGrid timeGrid)
: this(process, timeGrid, false)
{
}
public EqualJumpsBinomialTree(StochasticProcess1D process, double end, int steps)
: base(process, end, steps)
{
}
/*! Returns an approximation of the variance defined as
* \f$ \sigma(t_0, x_0)^2 \Delta t \f$. */
public double variance(StochasticProcess1D process, double t0, double x0, double dt)
{
double sigma = process.diffusion(t0, x0);
return(sigma * sigma * dt);
}
public ShortRateDynamics(StochasticProcess1D process)
{
process_ = process;
}
public AdditiveEQPBinomialTree(StochasticProcess1D process, double end, int steps, double strike)
: base(process, end, steps)
{
up_ = -0.5 * driftPerStep_ + 0.5 * Math.Sqrt(4.0 * process.variance(0.0, x0_, dt_) - 3.0 * driftPerStep_ * driftPerStep_);
}
public Trigeorgis factory(StochasticProcess1D process, double end, int steps, double strike)
{
return(new Trigeorgis(process, end, steps, strike));
}
/*! Returns an approximation of the drift defined as
\f$ \mu(t_0, x_0) \Delta t \f$. */
public double drift(StochasticProcess1D process, double t0, double x0, double dt)
{
return process.drift(t0, x0)*dt;
}
/*! Returns an approximation of the variance defined as
\f$ \sigma(t_0, x_0)^2 \Delta t \f$. */
public double variance(StochasticProcess1D process, double t0, double x0, double dt)
{
double sigma = process.diffusion(t0, x0);
return sigma*sigma*dt;
}
protected ShortRateDynamics(StochasticProcess1D process)
{
process_ = process;
}
/*! Returns an approximation of the diffusion defined as
\f$ \sigma(t_0, x_0) \sqrt{\Delta t} \f$. */
public double diffusion(StochasticProcess1D process, double t0, double x0, double dt)
{
return process.diffusion(t0, x0) * Math.Sqrt(dt);
}
public Joshi4 factory(StochasticProcess1D process, double end, int steps, double strike)
{
return(new Joshi4(process, end, steps, strike));
}
public LeisenReimer factory(StochasticProcess1D process, double end, int steps, double strike)
{
return(new LeisenReimer(process, end, steps, strike));
}
public Tian factory(StochasticProcess1D process, double end, int steps, double strike)
{
return(new Tian(process, end, steps, strike));
}
public JarrowRudd factory(StochasticProcess1D process, double end, int steps, double strike)
{
return(new JarrowRudd(process, end, steps, strike));
}
/*! Returns an approximation of the drift defined as
* \f$ \mu(t_0, x_0) \Delta t \f$. */
public double drift(StochasticProcess1D process, double t0, double x0, double dt)
{
return(process.drift(t0, x0) * dt);
}
public CoxRossRubinstein factory(StochasticProcess1D process, double end, int steps, double strike)
{
return(new CoxRossRubinstein(process, end, steps, strike));
}
/*! Returns an approximation of the diffusion defined as
* \f$ \sigma(t_0, x_0) \sqrt{\Delta t} \f$. */
public double diffusion(StochasticProcess1D process, double t0, double x0, double dt)
{
return(process.diffusion(t0, x0) * Math.Sqrt(dt));
}
public void testSingle(StochasticProcess1D process,
string tag,
bool brownianBridge,
double expected,
double antithetic)
{
ulong seed = 42;
double length = 10;
int timeSteps = 12;
var rsg = (InverseCumulativeRsg<RandomSequenceGenerator<MersenneTwisterUniformRng>
,InverseCumulativeNormal>)
new PseudoRandom().make_sequence_generator(timeSteps, seed);
PathGenerator<IRNG> generator = new PathGenerator<IRNG>(process,
length,
timeSteps,
rsg,
brownianBridge);
int i;
for (i=0; i<100; i++)
generator.next();
Sample<Path> sample = generator.next();
double calculated = sample.value.back();
double error = Math.Abs(calculated-expected);
double tolerance = 2.0e-8;
if (error > tolerance)
{
Assert.Fail("using " + tag + " process "
+ (brownianBridge ? "with " : "without ")
+ "brownian bridge:\n"
//+ std::setprecision(13)
+ " calculated: " + calculated + "\n"
+ " expected: " + expected + "\n"
+ " error: " + error + "\n"
+ " tolerance: " + tolerance);
}
sample = generator.antithetic();
calculated = sample.value.back();
error = Math.Abs(calculated-antithetic);
tolerance = 2.0e-7;
if (error > tolerance)
{
Assert.Fail("using " + tag + " process "
+ (brownianBridge ? "with " : "without ")
+ "brownian bridge:\n"
+ "antithetic sample:\n"
//+ setprecision(13)
+ " calculated: " + calculated + "\n"
+ " expected: " + antithetic + "\n"
+ " error: " + error + "\n"
+ " tolerance: " + tolerance);
}
}
public AdditiveEQPBinomialTree factory(StochasticProcess1D process, double end, int steps, double strike)
{
return(new AdditiveEQPBinomialTree(process, end, steps, strike));
}
