Calculates some basic statistical parameters of a distribution: sum, mean, variance, and provides the normal approximation. More...

#include <OpenMS/MATH/STATISTICS/BasicStatistics.h>

Public Types
typedef RealT	RealType
	The real type specified as template argument.
typedef std::vector< RealType >	probability_container
typedef std::vector< RealType >	coordinate_container
Public Member Functions
	BasicStatistics ()
	Default constructor.
	BasicStatistics (BasicStatistics const &arg)
	Copy constructor.
BasicStatistics &	operator= (BasicStatistics const &arg)
	Assignment.
void	clear ()
	Set sum, mean, and variance to zero.
template<typename ProbabilityIterator >
void	update (ProbabilityIterator probability_begin, ProbabilityIterator const probability_end)
	This does the actual calculation.
template<typename ProbabilityIterator , typename CoordinateIterator >
void	update (ProbabilityIterator const probability_begin, ProbabilityIterator const probability_end, CoordinateIterator const coordinate_begin)
	This does the actual calculation.
RealType	mean () const
	Returns the mean.
void	setMean (RealType const &mean)
RealType	variance () const
	Returns the variance.
void	setVariance (RealType const &variance)
RealType	sum () const
	Returns the sum.
void	setSum (RealType const &sum)
RealType	normalDensity_sqrt2pi (RealType coordinate) const
	Returns the density of the normal approximation at point, multiplied by sqrt( 2 * pi ). This saves a division operation compared to normalDensity().
RealType	normalDensity (RealType const coordinate) const
	See normalDensity_sqrt2pi(). Returns the density of the normal distribution at point.
void	normalApproximation (probability_container &probability)
	The argument `probability` is filled with values according to the normal approximation. Its `size()` is not changed. The approximation takes place at coordinate positions 0, 1, ..., size()-1.
void	normalApproximation (probability_container &probability, typename probability_container::size_type const size)
void	normalApproximation (probability_container &probability, coordinate_container const &coordinate)
	The argument probability is filled with values according to the normal approximation. The second argument coordinate contains the positions where the approximation takes place. probability.size() is set to coordinate.size().
Static Public Member Functions
static RealType	sqrt2pi ()
	Returns sqrt( 2 * pi ), which is useful to normalize the result of normalDensity_sqrt2pi().
Protected Attributes
Protected Members

RealType	mean_
RealType	variance_
RealType	sum_
Private Member Functions
Private Methods

void	normalApproximationHelper_ (probability_container &probability, typename probability_container::size_type const size)
void	normalApproximationHelper_ (probability_container &probability, coordinate_container const &coordinate)
Related Functions
(Note that these are not member functions.)
std::ostream &	operator<< (std::ostream &os, BasicStatistics &arg)
	A convenient overload for debugging purposes.

Detailed Description

template<typename RealT = DoubleReal>
class OpenMS::Math::BasicStatistics< RealT >

Calculates some basic statistical parameters of a distribution: sum, mean, variance, and provides the normal approximation.

The intended usage is as follows:

create an instance
set the basic statistical parameters by either
- calling one of the update() member functions, or
- using the set... methods
do something with the basic statistical parameters, e.g.
- using the get... methods, or
- obtain samples from a normal approximation with these parameters
- whatever member function you might want to add to this class ;-)

Member Typedef Documentation

typedef std::vector<RealType> coordinate_container

typedef std::vector<RealType> probability_container

typedef RealT RealType

The real type specified as template argument.

Reimplemented in AsymmetricStatistics< Real >.

Constructor & Destructor Documentation

BasicStatistics ( ) [inline]

Default constructor.

BasicStatistics ( BasicStatistics< RealT > const & arg ) [inline]

Copy constructor.

Member Function Documentation

void clear ( ) [inline]

Set sum, mean, and variance to zero.

RealType mean ( ) const [inline]

Returns the mean.

Referenced by ModelFitter< PeakType, FeatureType >::fitMZ_(), and ModelFitter< PeakType, FeatureType >::fitRT_().

void normalApproximation	(	probability_container &	probability,
		coordinate_container const &	coordinate
	)			`[inline]`

The argument probability is filled with values according to the normal approximation. The second argument coordinate contains the positions where the approximation takes place. probability.size() is set to coordinate.size().

void normalApproximation ( probability_container & probability ) [inline]

The argument probability is filled with values according to the normal approximation. Its size() is not changed. The approximation takes place at coordinate positions 0, 1, ..., size()-1.

void normalApproximation	(	probability_container &	probability,
		typename probability_container::size_type const	size
	)			`[inline]`

The argument probability is filled with values according to the normal approximation. Its size() is set to size. The approximation takes place at coordinate positions 0, 1, ..., size-1.

void normalApproximationHelper_	(	probability_container &	probability,
		typename probability_container::size_type const	size
	)			`[inline, private]`

void normalApproximationHelper_	(	probability_container &	probability,
		coordinate_container const &	coordinate
	)			`[inline, private]`

RealType normalDensity ( RealType const coordinate ) const [inline]

See normalDensity_sqrt2pi(). Returns the density of the normal distribution at point.

RealType normalDensity_sqrt2pi ( RealType coordinate ) const [inline]

Returns the density of the normal approximation at point, multiplied by sqrt( 2 * pi ). This saves a division operation compared to normalDensity().

BasicStatistics& operator= ( BasicStatistics< RealT > const & arg ) [inline]

Assignment.

void setMean ( RealType const & mean ) [inline]

void setSum ( RealType const & sum ) [inline]

void setVariance ( RealType const & variance ) [inline]

static RealType sqrt2pi ( ) [inline, static]

Returns sqrt( 2 * pi ), which is useful to normalize the result of normalDensity_sqrt2pi().

RealType sum ( ) const [inline]

Returns the sum.

void update	(	ProbabilityIterator	probability_begin,
		ProbabilityIterator const	probability_end
	)			`[inline]`

This does the actual calculation.

You can call this as often as you like, using different input vectors.

Referenced by ModelFitter< PeakType, FeatureType >::fit().

void update	(	ProbabilityIterator const	probability_begin,
		ProbabilityIterator const	probability_end,
		CoordinateIterator const	coordinate_begin
	)			`[inline]`