The Fib multimedia system
Fib is a system for storing multimedia data (like images or films).
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros
Public Member Functions
fib::algorithms::nCluster::iClusterFunction< tDataPoint > Class Template Reference

#include <iClusterFunction.h>

Inheritance diagram for fib::algorithms::nCluster::iClusterFunction< tDataPoint >:
Inheritance graph
[legend]

List of all members.

Public Member Functions

virtual ~iClusterFunction ()
virtual double operator() (const std::pair< tDataPoint, unsigned long > &dataClusterCenter, const std::pair< tDataPoint, unsigned long > &dataPoint) const =0

Detailed Description

template<class tDataPoint>
class fib::algorithms::nCluster::iClusterFunction< tDataPoint >

Definition at line 62 of file iClusterFunction.h.

Constructor & Destructor Documentation

template<class tDataPoint>
virtual fib::algorithms::nCluster::iClusterFunction< tDataPoint >::~iClusterFunction ( )
inlinevirtual

pure virtual destructor

Definition at line 68 of file iClusterFunction.h.

Member Function Documentation

template<class tDataPoint>
virtual double fib::algorithms::nCluster::iClusterFunction< tDataPoint >::operator() ( const std::pair< tDataPoint, unsigned long > &  dataClusterCenter,
const std::pair< tDataPoint, unsigned long > &  dataPoint 
) const
pure virtual

This operator should evalue a cluster value for two data points. The value should indicate how good it is to cluster the secound data point dataPoint to the first dataClusterCenter. If dataPoint should be clustered to dataClusterCenter the returned value should be greater 0, else it should be lower or equal 0. The function should have the following properties:

  • 0 < evalue( dataClusterCenter, dataPoint ) -> evalue( dataPoint, dataClusterCenter ) < 0
  • evalue( dataClusterCenter, dataPoint ) < 0 -> 0 < evalue( dataPoint, dataClusterCenter )
  • 0 < evalue( dataClusterCenter, dataPoint ) and 0 < evalue( dataPoint, data3 ) -> evalue( dataPoint, data3 ) < evalue( dataClusterCenter, data3 ) If some of theas conditions are not given, the clustering result may be not predictebel (undesired).
Parameters:
dataClusterCenterthe data point, to which to evalue the value that dataPoint is clustered to it
dataPointthe data point to cluster/subsume by dataClusterCenter
Returns:
the cluster value for dataClusterCenter and dataPoint
The documentation for this class was generated from the following file: