kml3d (version 0.7)

clustering: ~ Function: clustering ~

Description

clustering is the constructor of the class Clustering.

Usage

clustering(xLongData, yPartition, convergenceTime = 0, multiplicity = 1,criterionNames = c("calinski", "ray","davies","random"),
  algorithm = c(algo = "kmeans", startCond = "", imputation = "LI-Bissectrice")
)

Arguments

xLongData

[LongData]: longitudinal data on which the clusterization has been run.

yPartition

[Partition]: object that will be turn into a Clusterization

convergenceTime

[numeric]: if the clusterization has been obtained through an algorithm, save the number of steps of this algorithm before convergence.

multiplicity

[numeric]: if the clusterization has been obtained several time, this variable saves the number of time that this particular Clustering is obtained.

criterionNames

[vector(character)]: criterions used to evaluate the quality of the partitioning.

algorithm

[vector3(character)]: This variable hold informations about the algorithm used to get the Clustering. It has three value. The first (named algo) is the algorithm used ; the second (named startCond

Value

• Object of class Clustering.

Author(s)

Christophe Genolini INSERM U669 / PSIGIAM: Paris Sud Innovation Group in Adolescent Mental Health Modal'X / Universite Paris Ouest-Nanterre- La Defense Contact author : [e-mail]

Details

In Kml3d, strickly speaking, a Partition is just a sequence of letters (independent of any trajectories) ; a Clustering is a Partition associated with a set of trajectories with some additional information like qualities criterion, size of the clusters, algorithm used to get the Clustering...

References

Article "KmL: K-means for Longitudinal Data", in Computational Statistics, Volume 25, Issue 2 (2010), Page 317. Web site: http://christophe.genolini.free.fr/kml

Examples

### Creation of a partition
part <- partition(rep(c(1,2),75),2)

### Some trajectories
myLd <- gald(clusterLD=FALSE)

### Tranformation of part into a Clusterization
clustering(myLd,part)