[R-sig-eco] Re : Problem with PCA

[momadou sow]

Hi again, 
I try to improve my PCA it is the first once I do it.  Many variables are much correlated (please see the attached file) and thus the labels became unreadable. I increased and decreased the size of labels, but it is always unreadable. 
Is there a way to solve this problem.
Thank you in advance 
Komine
 

 
De : florian gollnow <florian.gollnow at gmail.com>
À : 'momadou sow' <momadou at yahoo.fr>; R-sig-ecology at r-project.org
Envoyé le : Lundi 15 Août 2011 15h59
Objet : RE: [R-sig-eco] Problem with PCA

You might like to try:
zouk<-read.table("C:\\Users\\Tableau1.csv",sep=";",dec=",",header=TRUE,row.n
ames=1)
zouk
TableauS1<-na.omit(zouk)      <----- CHANGED
library(FactoMineR)
res.pca<-PCA(TableauS1)

hope that helps
florian


-----Original Message-----
From: r-sig-ecology-bounces at r-project.org
[mailto:r-sig-ecology-bounces at r-project.org] On Behalf Of momadou sow
Sent: 15 August 2011 14:34
To: R-sig-ecology at r-project.org
Subject: [R-sig-eco] Problem with PCA

Hi,
I try to do
PCA, but I have some problems. Here a sample of my data in the attached
file.
I used this
code:
 
zouk<-read.table("C:\\Users\\Tableau1.csv",sep=";",dec=",",header=TRUE,row.n
ames=1)
zouk
TableauS1<-na.omit("C:\\Users\\Tableau1.csv")
# to delete NA
library(FactoMineR)
res.pca<-PCA(TableauS1)
But there a
problem because R give this message: The following variables are not
quantitative:  X
Thus, I
used: is.numeric(zouk$ARVI) to verified if each variable is numeric. They
are numeric. 
Could you
help me to see my problem? 
Thank you
in advance
Komine