Inheritance diagram for npstat::GriddedRobustRegressionStop:

Public Member Functions
	GriddedRobustRegressionStop (unsigned long maxcalls, double lossImprovementTarget=-1.0e300, double lossValueTarget=-1.0e300)

virtual bool	operator() (const LocalLoss &lastLoss, const unsigned long &nReplacements) const

Additional Inherited Members
Public Types inherited from npstat::Functor2< bool, LocalLoss, unsigned long >
typedef bool	result_type

typedef LocalLoss	first_argument_type

typedef unsigned long	second_argument_type

Detailed Description

A simple stopping functor for iterative local gridded robust regression

Constructor & Destructor Documentation

npstat::GriddedRobustRegressionStop::GriddedRobustRegressionStop	(	unsigned long	maxcalls,
		double	lossImprovementTarget = `-1.0e300`,
		double	lossValueTarget = `-1.0e300`
	)

inlineexplicit

The robust regression will be stopped when at least one of the following conditions is satisfied:

– Number of iterations reaches "maxcalls"

– The loss function improvement due to point replacement operation becomes smaller than "lossImprovementTarget"

– The loss value itself becomes smaller than "lossValueTarget".

The documentation for this class was generated from the following file: