GeoXp (version 1.6.2)

pcamap: Generalized Principal Component Analysis and map

Description

The function pcamap() draws the plots summarizing a generalized Principal Component Analysis (PCA), made with genpca. It draws the scatterplot of the individuals projected on a chosen principal component plane (with their percentage of inertia), together with the scatterplot of the variables projected into the same plane with the quality of representation in order to interpret the principal component axes. The individuals scatterplot interacts with the map.

Usage

pcamap(sp.obj, names.var, direct=c(1,2),
weight=rep(1/nrow(sp.obj),length=nrow(sp.obj)), metric=diag(length(names.var)), 
center=NULL, reduce=TRUE, qualproj=FALSE, names.attr=names(sp.obj), criteria=NULL, 
carte=NULL, identify=FALSE, cex.lab=0.8, pch=16, col="lightblue3",
xlab=paste(direct[1]), ylab=paste(direct[2]), axes=FALSE, lablong="", lablat="")

Arguments

sp.obj

object of class extending Spatial-class

names.var

a vector of character; attribute names or column numbers in attribute table

direct

Two-dimension vector containing the numbers of principal axes to plot

weight

vector of size n of weight (by default : weight=t(1/n,...,1/n))

metric

matrix $p x p$ (by default : metric=Identity matrix)

center

A vector of size p of the gravity center which by default is equal to $bar(X)$

reduce

if TRUE, reduced PCA

qualproj

if TRUE, print the quality of representation of individuals

names.attr

names to use in panel (if different from the names of variable used in sp.obj)

criteria

a vector of boolean of size the number of Spatial unit, which permit to represent preselected sites with a cross, using the tcltk window

carte

matrix with 2 columns for drawing spatial polygonal contours : x and y coordinates of the vertices of the polygon

identify

if not FALSE, identify plotted objects (currently only working for points plots). Labels for identification are the row.names of the attribute table row.names(as.data.frame(sp.obj)).

cex.lab

character size of label

pch

a vector of symbol which must be equal to the number of group else all sites are printed in pch[1]

col

a vector of colors which must be equal to the number of group else all sites and all bars are printed in col[1]

xlab

a title for the graphic x-axis

ylab

a title for the graphic y-axis

axes

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

Value

In the case where user click on save results button, a list is created as a global variable in last.select object. obs, corresponds to the number of spatial units selected just before leaving the Tk window, inertia vector of size p with percent of inertia of each component, casecoord matrix $n x p$ of individuals, varcoord matrix $n x p$ of principal components.

Details

Let $$D=diag(\lambda_1,...,\lambda_p)$$ $$1_p=(1,...,1)'$$ Let the coordinates of individuals in the principals components $$CC=(C_1',...,C_n')'$$ with $C_i=(C_i^1,...,C_i^p)$ Let the coordinates of variables in the principals components $$CC=(V_1',...,V_p')'$$ with $V_i=(V_i^1,...,V_i^p)$ Part of inertia : $$(\frac{\lambda_1}{\sum_i\lambda_i},...,\frac{\lambda_p}{\sum_i\lambda_i})'$$ Quality of representation of individual k projected on plane (i,j): $$Qu=\sqrt{\frac{(C_k^i)^2+(C_k^j)^2}{\sum_l(C_k^l)^2}}$$ Quality of representation of variable k projected on plane (i,j): $$VQu=\sqrt{\frac{(V_k^i)^2+(V_k^j)^2}{\sum_l(V_k^l)^2}}$$

References

Thibault Laurent, Anne Ruiz-Gazen, Christine Thomas-Agnan (2012), GeoXp: An R Package for Exploratory Spatial Data Analysis. Journal of Statistical Software, 47(2), 1-23.

Caussinus H., Fekri M., Hakam S., Ruiz-Gazen A. (2003) , A monitoring display of Multivariate Outliers, Computational Statistics and Data Analysis, vol. 44, 1-2, 237-252.

See Also

clustermap,genpca

Examples

###
# Data Colombus
require("maptools")
require("robustbase")
example(columbus)

# a basic PCA on 7 variables
pcamap(columbus,c(6:12), qualproj=TRUE, identify=TRUE)

###
# data boston
data(boston)

# SpatialPoints object
boston.sp<-SpatialPoints(cbind(boston.utm[,1],boston.utm[,2]))

# SpatialPointsDataFrame object
boston.spdf <- SpatialPointsDataFrame(boston.sp, boston.c)

# a basic PCA on 7 variables
pcamap(boston.spdf, c(7:8,10,12:15), identify=TRUE, cex.lab=0.5)

# generalized PCA : user have to construct a new metric and a vector
# of gravity center, by using for exampe covMcd
cov.boston<-covMcd(boston.c[,c(7:8,10,12:15)],alpha=.75)
b.center<-cov.boston$center
b.cov<-cov.boston$cov

# example of use of pcamap
pcamap(boston.spdf, c(7:8,10,12:15), metric=b.cov, center=b.center,identify=TRUE,
cex.lab=0.5)