[R] map points from scatterplot3d onto 2d fitted plane

[Uwe Ligges]

Jacqueline Hall wrote:
> Dear R helpers,
> 
> I have a 3D scatter plot that I have generated from scatterplot3d (which
> looks great- thanks!) and I can see that the points in my graph fall in a
> plane.
> Following the example 5 from 3D scatter plot (below) I have fitted a
> regression plane.
> Now what I would like to do is a rotation  so that my new co-ordinate system
> is about the fitted plane (by finding the normal to the plane using
> crossprod?) and then map the data points onto this plane.
> 
> I played with the function xyz.convert from s3d but i think this is mapping
> to the xy plane whereas i would like to map onto my fitted plane.
> 
> Is there a package/function that can help me do this that I have missed? or
> does anyone have any suggestions / sample code to get me started in the
> right direction?


Typing

   vignette(package="scatterplot3d")

shows the package comes with a package vignette. Further on typing

   vignette("s3d", package="scatterplot3d")

shows the vignette which has been published (as a former more compact 
version) as

  Ligges, U. and Mächler, M. (2003). Scatterplot3d - an R Package for 
Visualizing Multivariate Data. Journal of Statistical Software 8(11), 1-20.


See page 22 of the package vignette for an example that could be 
rewritten for the trees data you used below as follows:

   s3d <- scatterplot3d(trees, angle=55, scale.y=0.7, pch=16,
                        main="scatterplot3d - 5")
   my.lm <- lm(Volume ~ Girth + Height, data = trees)
   s3d$plane3d(my.lm, lty.box = "solid")
   orig <- s3d$xyz.convert(trees)
   plane <- s3d$xyz.convert(trees[,1], trees[,2], fitted(my.lm))
   i.negpos <- 1 + (resid(my.lm) > 0)
   segments(orig$x, orig$y, plane$x, plane$y,
            col = c("blue", "red")[i.negpos], lty = (2:1)[i.negpos])


Best,
Uwe Ligges




> Thanks very much for your help!
> Jacqueline
> 
> #R-2.6
> ## example 5
>        data(trees)
>        s3d <- scatterplot3d(trees, type="h", highlight.3d=TRUE,
>            angle=55, scale.y=0.7, pch=16, main="scatterplot3d - 5")
>        # Now adding a regression plane to the "scatterplot3d"
>        attach(trees)
>        my.lm <- lm(Volume ~ Girth + Height)
>        s3d$plane3d(my.lm, lty.box = "solid")
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.