#include <DiscreteDistributions1D.hh>

Inheritance diagram for npstat::Poisson1D:

Public Member Functions
	Poisson1D (double lambda)

	Poisson1D (const Poisson1D &)

Poisson1D &	operator= (const Poisson1D &)

virtual Poisson1D *	clone () const

double	mean () const

double	probability (long x) const

double	cdf (double x) const

double	exceedance (double x) const

long	quantile (double x) const

unsigned	random (AbsRandomGenerator &g, long *generatedRandom) const

virtual gs::ClassId	classId () const

virtual bool	write (std::ostream &os) const

Public Member Functions inherited from npstat::AbsDiscreteDistribution1D
bool	operator== (const AbsDiscreteDistribution1D &r) const

bool	operator!= (const AbsDiscreteDistribution1D &r) const

Static Public Member Functions
static const char *	classname ()

static unsigned	version ()

static Poisson1D *	read (const gs::ClassId &id, std::istream &in)

Static Public Member Functions inherited from npstat::AbsDiscreteDistribution1D
static const char *	classname ()

static unsigned	version ()

static AbsDiscreteDistribution1D *	read (const gs::ClassId &id, std::istream &is)

Protected Member Functions
virtual bool	isEqual (const AbsDiscreteDistribution1D &o) const

Detailed Description

The Poisson distribution. The calculations of cdf and quantiles, together with generation of random numbers, are implemented by constructing the cdf lookup table. Once such a table is constructed, calculations of this kind become pretty fast. Nevertheless, since construction of the complete cdf table takes some time (especially for large Poisson lambdas), one can imagine situations in which it may be more efficient to set up a Poisson process instead. For example, if you need to generate a few random numbers for many different values of lambda, you might be able to save some CPU time by following the Poisson process route.