ConvergenceStatistics< T, U > Class Template Reference

statistics class with convergence table More...

#include <ql/math/statistics/convergencestatistics.hpp>

Inherits T.

Public Types
typedef T::value_type	value_type
typedef std::vector< std::pair< Size, value_type > >	table_type

Public Member Functions
	ConvergenceStatistics (const T &stats, const U &rule=U())
	ConvergenceStatistics (const U &rule=U())
void	add (const value_type &value, Real weight=1.0)
template<class DataIterator >
void	addSequence (DataIterator begin, DataIterator end)
template<class DataIterator , class WeightIterator >
void	addSequence (DataIterator begin, DataIterator end, WeightIterator wbegin)
void	reset ()
const std::vector< std::pair< Size, value_type > > &	convergenceTable () const

Detailed Description

template<class T, class U = DoublingConvergenceSteps>
class QuantLib::ConvergenceStatistics< T, U >

statistics class with convergence table

This class decorates another statistics class adding a convergence table calculation. The table tracks the convergence of the mean.

It is possible to specify the number of samples at which the mean should be stored by mean of the second template parameter; the default is to store \( 2^{n-1} \) samples at the \( n \)-th step. Any passed class must implement the following interface:

Size initialSamples() const;
Size nextSamples(Size currentSamples) const;

as well as a copy constructor.

Tests:

results are tested against known good values.