fib::algorithms::nCluster Namespace Reference

Classes
class	cClusterFunArea
class	cClusterFunExp
class	iClusterFunction
class	iDistFunction

Functions
template<class tDataPoint >
map< tDataPoint, tDataPoint >	cluster (const map< tDataPoint, unsigned long > &mapData, const iClusterFunction< tDataPoint > &evaluePoints)

---

Function Documentation

template<class tDataPoint >

map< tDataPoint, tDataPoint > fib::algorithms::nCluster::cluster	(	const map< tDataPoint, unsigned long > &	mapData,
		const iClusterFunction< tDataPoint > &	evaluePoints
	)

This function clusters the given data. This will be done with an given formular evaluePoints. An data point D can be in the cluster to an other data point C, if 0 < evaluePoints.evalue( C, D ) If ther are more data points C_i for which the formulare is fulfilled, the data point will be mapped to the data point wher evaluePoints( C_i, D ) is the maximal.

Parameters:

mapData	the data to cluster key: the datapoint value: the number of elements for the data point
evaluePoints	the class for evaluing the cluster value for two data points

Returns:

a map with: -key: the data points of the given mapData -value: the data point, which is the center of the cluster, in which the key data point is; so the key data point should be maped to the value data point