RansacModel< ModelT > Class Template Reference

Generic plug-in template base class using 'Curiously recurring template pattern' (CRTP) to allow for arbitrary RANSAC models (e.g. linear or quadratic fits). More...

#include <OpenMS/MATH/MISC/RANSACModel.h>

Public Types
typedef std::pair< double, double >	DPair
typedef std::vector< DPair >	DVec
typedef DVec::const_iterator	DVecIt
typedef std::vector< double >	ModelParameters

Public Member Functions
ModelParameters	rm_fit (const DVecIt &begin, const DVecIt &end) const
	fit a model and return its parameters More...
double	rm_rsq (const DVecIt &begin, const DVecIt &end) const
	Returns the R-squared of the data applied to the model (computed on-the-fly). More...
double	rm_rss (const DVecIt &begin, const DVecIt &end, const ModelParameters &coefficients) const
	calculates the residual sum of squares of the input points according to the model More...
DVec	rm_inliers (const DVecIt &begin, const DVecIt &end, const ModelParameters &coefficients, double max_threshold) const

Detailed Description

template<class ModelT = int>
class OpenMS::Math::RansacModel< ModelT >

Generic plug-in template base class using 'Curiously recurring template pattern' (CRTP) to allow for arbitrary RANSAC models (e.g. linear or quadratic fits).

An actual model should derive from this class using CRTP (see RansacModelLinear for an example) and implement the *_impl() functions.

Member Typedef Documentation

DPair

typedef std::pair<double, double> DPair

DVec

typedef std::vector<DPair> DVec

DVecIt

typedef DVec::const_iterator DVecIt

ModelParameters

typedef std::vector<double> ModelParameters

Member Function Documentation

rm_fit()

ModelParameters rm_fit ( const DVecIt &  begin, const DVecIt &  end  ) const

inline

fit a model and return its parameters

rm_inliers()

DVec rm_inliers ( const DVecIt &  begin, const DVecIt &  end, const ModelParameters &  coefficients, double  max_threshold  ) const

inline

calculates the squared residual of each input point vs. the model. All points with an error <= max_threshold, are accepted as inliers and returned.

rm_rsq()

double rm_rsq ( const DVecIt &  begin, const DVecIt &  end  ) const

inline

Returns the R-squared of the data applied to the model (computed on-the-fly).

Takes as input a standard vector of a standard pair of points in a 2D space.

Parameters

begin	Iterator to first pair
end	Past-end iterator to last pair

Returns

: R-squared value

rm_rss()

double rm_rss ( const DVecIt &  begin, const DVecIt &  end, const ModelParameters &  coefficients  ) const

inline

calculates the residual sum of squares of the input points according to the model