[R-sig-Geo] matching shading and contour

F. De Sales fhds2000 at yahoo.com
Sun Apr 27 05:34:58 CEST 2008


   Dear Frede,
   Thank you very much for helping me with this
matter.
I really appreciate you taking the time to do this.
   I tried your suggestion but I got the following
error message after entering the interp command:

P.interp <- interp( xx,yy,P)

Error in interp.old(x, y, z, xo = xo, yo = yo, ncp =
0, extrap = extrap,  : 
  Lengths of x, y, and z do not match

   Do you know why?  I am kind of a novice to R. 
   Thanks again

   F. DeSales
 
   
--- Frede Aakmann Tøgersen <FredeA.Togersen at agrsci.dk>
wrote:

> Thank you for finally (you've asked at the r-help
> list too) for providing a minimal example. It makes
> it easier for us to help you.
>  
> Try using the interp() function from the "akima"
> package. E.g.
> 
> install.packages("akima")
> library(akima)
>  
> dim(P)
>  
> P.interp <- interp(xyz$x, xyz$y, xyz$z)
>  
> str(P.interp)
>  
> mynl <- 20
>
filled.contour(P.interp$x,P.interp$y,P.interp$z,ylim=c(2,5),
>    zlim=c(-1.5,1.5),nlevels=mynl,
>    color=cm.colors,
>    plot.axes = {
>      
> contour(P.interp$x,P.interp$y,P.interp$z,add=T,
> col="grey", ylim=c(2,5),
>       zlim=c(-1.5,1.5),nlevels=mynl,
>       drawlabels=F)
>       axis(1,1:length(xx))
>       axis(2,1:length(yy))
>     })
>  
> You can play with the settings of interp() together
> with the settings of nlevels of filled.contour() in
> order to get something close to what you want:
> 
> P.interp <- interp(xyz$x, xyz$y, xyz$z,
>                    xo = seq(min(xyz$x), max(xyz$x),
> length = 100),
>                    yo = seq(min(xyz$y), max(xyz$y),
> length = 100))
>  
> mynl <- 40, # or 10, 20, 50, .......
>
filled.contour(P.interp$x,P.interp$y,P.interp$z,ylim=c(2,5),
>    zlim=c(-1.5,1.5),nlevels=mynl,
>    color=cm.colors,
>    plot.axes = {
>      
> contour(P.interp$x,P.interp$y,P.interp$z,add=T,
> col="grey", ylim=c(2,5),
>       zlim=c(-1.5,1.5),nlevels=mynl,
>       drawlabels=F)
>       axis(1,1:length(xx))
>       axis(2,1:length(yy))
>     })
> 
> regards Frede Aakman Tøgersen
> 
> ________________________________
> 
> Fra: r-sig-geo-bounces at stat.math.ethz.ch på vegne af
> F. De Sales
> Sendt: fr 25-04-2008 21:08
> Til: r-sig-geo at stat.math.ethz.ch
> Emne: [R-sig-Geo] matching shading and contour
> 
> 
> 
>     Hello everyone.
>     Once again I need you help. How to make sure
> that
> contour lines and shading match when using the
> command
> filled.contour?
>     This is my example:
> 
> P = array(0,c(10,5))
> P [1,]= c(-0.0708, -0.09906,  0.2346,  0.1556,
> -0.1136)
> P [2,]= c(-0.1086, -0.07813,  0.4122,  0.3290,
> -0.2537)
> P [3,]= c(-0.0661, -0.16677,  0.5109,  0.6276,
> -0.4118)
> P [4,]= c(-0.0728, -0.05032,  0.2441,  0.7046,
> -0.8731)
> P [5,]= c(-0.0940,  0.00464,  0.0212,  0.3605,
> -1.2655)
> P [6,]= c(-0.2043, -0.18422, -0.1460,  0.0917,
> -1.0035)
> P [7,]= c(-0.2263, -0.17598,  0.0318, -0.0685,
> -0.8082)
> P [8,]= c(-0.3277, -0.16696,  0.2180, -0.0898,
> -0.6687)
> P [9,]= c(-0.3139, -0.19312,  0.4357, -0.1240,
> -0.3028)
> P[10,]= c(-0.1368,  0.15187,  0.1775,  0.0379,
> -0.0158)
> xx = seq(1,10)
> yy = seq(1,5)
> filled.contour(xx,yy,P,ylim=c(2,5),
>    zlim=c(-1.5,1.5),nlevels=20,
>    color=cm.colors,
>    plot.axes = {
>       contour(xx,yy,P,add=T, col="grey",
> ylim=c(2,5),
>       zlim=c(-1.5,1.5),nlevels=20,
>       drawlabels=F)
>       axis(1,1:length(xx))
>       axis(2,1:length(yy))
>     })
> 
>     You can see that in the center of the plot the
> contour lines do no match the shaded pink area.
> 
>     Any idea?
>     Thank you again for your help.
>     F. De Sales
> 
> 
>      
>
____________________________________________________________________________________
> 
> [[elided Yahoo spam]]
> 
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
> 
> 
> 



      ____________________________________________________________________________________

[[elided Yahoo spam]]




More information about the R-sig-Geo mailing list