getMatDist: Computation of a dissimilarity matrix

In MathieuEmily/SpatialClustering: Detection of clusters of bi-dimensional points

Description

getMatDist allows the dissimilarity matrix between the 2-dimensional points of the studied dataset.

Usage

	getMatDist(Proba,Traj)

Arguments

Proba	A list of n matrices of pairwise probabilities (each of size (n-1)x(n-1)). Typically the Probabilities object obtain from generateListTandP.
Traj	A list of n list of trajectories. Typically the Trajectories object obtain from generateListTandP.

Details

getMatDist proposes the implementation of the minimum of all pairwise probabilities. It relates to the single-linkage criterion in hierarchical clustering. More details are given in Bar-Hen et al. (2015).

Value

An nxn matrix containing all pairwise dissimilarity between the n points of the dataset.

References

A. Bar-Hen, M. Emily and N. Picard. (2015) Spatial Cluster Detection Using Nearest Neighbour Distance, Spatial Statistics, Vol. 14, pages 400-411.

See Also

generateListTandP, getClusters, SpatialClustering

Examples

## Example of a study of tree location 
data(dataExample)


## Extraction of the data and the window
dDicor <- dataExample$data
w0 <- dataExample$w0

## the Homogeneous case

List.Dicor.H <- generateListTandP(dDicor,w0,Homogeneous=TRUE)
MatDist.Dicor.H <- getMatDist(List.Dicor.H$Probabilities,List.Dicor.H$Trajectories) 

## the Inhomogeneous case

## Extraction of the covariate
Z.Pente <- dataExample$Z.Pente

List.Dicor.I <- generateListTandP(dDicor,w0=w0,Homogeneous=FALSE,Z=Z.Pente)
MatDist.Dicor.I <- getMatDist(List.Dicor.I$Probabilities,List.Dicor.I$Trajectories)

MathieuEmily/SpatialClustering documentation built on May 7, 2019, 4:33 p.m.