[R] Installing rgl

[Milan Bouchet-Valat]

Le mercredi 22 février 2012 à 04:51 -0800, aoife a écrit :
> Because for example I've done this:
> 
> I made a simple table:
> 
> 	A	B	C
> G1	1	34	231
> G2	231	1	0.1
> G3	12	0.02	23
> 
> 
> and ran a correspondance analysis on it using the ca package:
> 
> > library(ca)
> Loading required package: rgl
> Warning messages:
> 1: In rgl.init(initValue) : RGL: unable to open X11 display
> 2: In fun(libname, pkgname) : error in rgl_init
> > ca(table)
> 
>  Principal inertias (eigenvalues):
>            1        2       
> Value      0.924588 0.007655
> Percentage 99.18%   0.82%   
> 
> 
>  Rows:
>                G1        G2        G3
> Mass     0.498950  0.435362  0.065689
> ChiDist  0.911602  1.079139  0.401921
> Inertia  0.414636  0.506996  0.010611
> Dim. 1  -0.947585  1.122145 -0.239642
> Dim. 2  -0.326025 -0.194244  3.763758
> 
> 
>  Columns:
>                 A         B         C
> Mass     0.457683  0.065689  0.476628
> ChiDist  1.046690  0.943948  0.883798
> Inertia  0.501419  0.058531  0.372293
> Dim. 1   1.088538 -0.923587 -0.917983
> Dim. 2  -0.001466 -3.656541  0.505351
> 
> 
> But i'm having trouble actually visualising a graph? Does anyone know, is it
> because of the rgl issue/ is there an alternative way to turn this
> correspondance analysis results into a graph? Basically i just want to
> picture this correspondance analysis results graphically, using any method
> :(
You only need rgl if you want to make a 3D plot of your CA. In most
cases, that's not required. So just try:
plot(ca)


Cheers