GClasses::GRowDistance Class Reference

This uses a combination of Euclidean distance for continuous attributes, and Hamming distance for nominal attributes. In particular, for each attribute, it calculates pA[i]-pB[i], squares it and takes the square root of that sum. For nominal attributes pA[i]-pB[i] is 0 if they are the same and 1 if they are different. More...

#include <GDistance.h>

Inheritance diagram for GClasses::GRowDistance:

Public Member Functions
	GRowDistance ()
	GRowDistance (GDomNode *pNode)
virtual	~GRowDistance ()
virtual void	init (const GRelation *pRelation, bool own)
	See the comment for GDistanceMetric::init. More...
virtual GDomNode *	serialize (GDom *pDoc) const
	See the comment for GDistanceMetric::serialize. More...
void	setDiffWithUnknown (double d)
	Specify the difference to use when one or more of the values is unknown. (If your data contains unknown values, you may want to normalize the known values to fall within some pre-determined range, so that it will be possible to select a reasonable value for this purpose.) More...
virtual double	squaredDistance (const double *pA, const double *pB) const
	Returns the distance between pA and pB. More...
Public Member Functions inherited from GClasses::GDistanceMetric
	GDistanceMetric ()
	GDistanceMetric (GDomNode *pNode)
virtual	~GDistanceMetric ()
double	operator() (const std::vector< double > &a, const std::vector< double > &b) const
	Return squaredDistance(a,b). Allows dissimilarity metrics to be used as function objects. Do not override. Override squaredDistance(a,b) instead. See GDistanceMetric::squaredDistance(const std::vector<double>&, const std::vector<double>&) More...
double	operator() (const double *pA, const double *pB) const
	Return squaredDistance(pA, pB). Allows dissimilarity metrics to be used as function objects. Do not override. Override squaredDistance(pA,pB) instead. See GDistanceMetric::squaredDistance(const double*, const double*) More...
const GRelation *	relation () const
	Returns the relation that specifies the meaning of the vector elements. More...
virtual double *	scaleFactors ()
	Returns a pointer to the vector of scale factors. This may be NULL if the metric does not use scale factors. More...
virtual double	squaredDistance (const std::vector< double > &a, const std::vector< double > &b) const
	Return the squared distance (or squared dissimilarity) between the two specified vectors. More...

Protected Attributes
double	m_diffWithUnknown
Protected Attributes inherited from GClasses::GDistanceMetric
bool	m_ownRelation
const GRelation *	m_pRelation

Additional Inherited Members
Static Public Member Functions inherited from GClasses::GDistanceMetric
static GDistanceMetric *	deserialize (GDomNode *pNode)
	Deserializes a distance metric. More...
Protected Member Functions inherited from GClasses::GDistanceMetric
GDomNode *	baseDomNode (GDom *pDoc, const char *szClassName) const
void	setRelation (const GRelation *pRelation, bool own)
	Sets the relation to use with this metric. Takes ownership of the relation iff own is true. More...

Detailed Description

This uses a combination of Euclidean distance for continuous attributes, and Hamming distance for nominal attributes. In particular, for each attribute, it calculates pA[i]-pB[i], squares it and takes the square root of that sum. For nominal attributes pA[i]-pB[i] is 0 if they are the same and 1 if they are different.

Constructor & Destructor Documentation

GClasses::GRowDistance::GRowDistance

(

)

GClasses::GRowDistance::GRowDistance

(

GDomNode *

pNode

)

virtual GClasses::GRowDistance::~GRowDistance ( )

inlinevirtual

Member Function Documentation

virtual void GClasses::GRowDistance::init ( const GRelation *  pRelation, bool  own  )

virtual

See the comment for GDistanceMetric::init.

Implements GClasses::GDistanceMetric.

virtual GDomNode* GClasses::GRowDistance::serialize ( GDom *  pDoc ) const

virtual

See the comment for GDistanceMetric::serialize.

Implements GClasses::GDistanceMetric.

void GClasses::GRowDistance::setDiffWithUnknown ( double  d )

inline

Specify the difference to use when one or more of the values is unknown. (If your data contains unknown values, you may want to normalize the known values to fall within some pre-determined range, so that it will be possible to select a reasonable value for this purpose.)

virtual double GClasses::GRowDistance::squaredDistance ( const double *  pA, const double *  pB  ) const

virtual

Returns the distance between pA and pB.

Implements GClasses::GDistanceMetric.

Member Data Documentation

double GClasses::GRowDistance::m_diffWithUnknown

protected