clustermap {GeoXp}R Documentation

Classification of dataset using kmeans.r or hclust.r and representation of clusters on a map.

Description

The function `clustermap' performs a classification of the sites from the variables included in $dataset$ and computes a bar plot of the clusters calculated. Classification methods comes from hclust.R (hierarchical cluster analysis) and kmeans.R (k-means clustering) and number of class is chosen with $clustnum$.

Usage

clustermap(long,lat,dataset,clustnum,method="kmeans",
           type=NULL,centers=NULL,scale=FALSE,listvar=NULL, listnomvar=NULL,
           carte=NULL,criteria=NULL,labvar="Cluster",
           label="",symbol=0,color=1,labmod="",
           axis=FALSE,lablong="", lablat="")

Arguments

long a vector $x$ of size $n$
lat a vector $y$ of size $n$
dataset matrix of variables of numeric values
clustnum Number of clusters
method two methos : `kmeans' by default or `hclust'
type If method=`hclust'', type=``complete'' by default (the possibilities are given in help(hclust) as `ward', `single', etc). If method=`kmeans', type="Hartigan-Wong" by default (the possibilities are given in help(kmeans) as `Forgy', etc)
centers If method='kmeans', user can give a matrix with initial cluster centers.
scale If scale=TRUE, the dataset is reducted.
listvar matrix of variables
listnomvar names of variables $listvar$
criteria a vector of size $n$ of boolean with TRUE on specific sites (these for non interactive selection)
carte matrix with 2 columns for drawing spatial polygonal contours : $x$ and $y$ coordinates of the vertices of the polygon
labvar name of variable $var$
label vector of size $n$ with names of each site
color 0 or 1 (by default), choice of representation of selected points. If 0, sites are represented in blue, if 1, sites are represented with different colors for each cluster
symbol 0 (by default) or 1, choice of representation of selected points. If 0, selected points are circles, if 1, selected points are stars
labmod names of Clusters
axis a boolean with TRUE for drawing axes on the map
lablong name of the x-axis that will be printed on the map
lablat name of the y-axis that will be printed on the map

Details

The two windows are interactive : the sites selected by a bar chosen on the bar plot are represented on the map in red and the values of sites selected on the map by `points' or `polygon' are represented in red on the bar plot. The dendogram is also drawn for 'hclust' method. In option, possibility to choose the classification method.

Value

A vector of boolean of size $n$ (TRUE if the site was in the last selection) and a vector of size $n$ whith the number of cluster for each site

Note

To use the functions hclust.r and kmeans.r, we take many arguments by default. If the user would like to modify these arguments, he should call these functions first and then use the function barmap.r to visualize the calculated clusters.

Author(s)

Thomas-Agnan C., Aragon Y., Ruiz-Gazen A., Laurent T., Robidou L.

References

Aragon Yves, Perrin Olivier, Ruiz-Gazen Anne, Thomas-Agnan Christine (2006), ``Statistique et Econométrie pour données géoréférencées : modèles et études de cas''

Murtagh, F (1985). ``Multidimensional Clustering Algorithms''.

Hartigan, J. A. and Wong, M. A. (1979). A K-means clustering algorithm. Applied Statistics 28, 100-108

See Also

barmap, pcamap

Examples

# données oldcol
data(oldcol)
data(oldcol.polys)
contours.OLD<-polylist2list(oldcol.polys)
obs<-clustermap(oldcol$X, oldcol$Y,oldcol[,7:12],3,listvar=oldcol,
listnomvar=names(oldcol),carte=contours.OLD,criteria=(oldcol$CP==1))

# data avifaune 
#library(ade4)
#data(mafragh)
#d<-distchi2(mafragh$flo)

#obs<-clustermap(x<-mafragh$xy[,1],y<-mafragh$xy[,2],d,
#4,method="hclust",type="ward",
#listvar=mafragh$partition,listnomvar="class")

#clas<-obs$vectclass
#e<-by(d,clas,mean)
#g<-rbind(e[1]$'1',e[2]$'2',e[3]$'3',e[4]$'4')

#clustermap(x<-mafragh$xy[,1],y<-mafragh$xy[,2],
#d,4,centers=g,listvar=mafragh$partition,listnomvar="class")

[Package GeoXp version 1.0 Index]