[R] OT: 3d surfaces with transparency

[Deepayan Sarkar]

On 12/8/07, Ben Bolker <bolker at ufl.edu> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Deepayan Sarkar wrote:
> > On 12/8/07, Ben Bolker <bolker at ufl.edu> wrote:
> >> -----BEGIN PGP SIGNED MESSAGE-----
> >> Hash: SHA1
> >>
> >>
> >>   I would be grateful if anyone had suggestions
> >> about software that could (1) create 3D surface
> >> plots, (2) handle transparency/alpha blending,
> >> (3) generate output in some vector graphics format
> >> that preserved the transparency.  I could also
> >> live with a combination of two programs, one
> >> to generate the basic figure and another to
> >> modify the output surface to a transparent
> >> color (but preserving vector-ness).
> >>
>  [snip]
> >>   Any ideas???
> >
> > You could consider wireframe from lattice, but this has many caveats.
> > For an example, see
> >
> > http://dsarkar.fhcrc.org/lattice/book/figures.html?chapter=06;figure=06_18;theme=stdColor;code=right
> >
> > For that matter, whats wrong with persp?
> >
> > -Deepayan
>
>
>   I hadn't thought about the fact that transparency is easier
> than it used to be (esp. with cairo device/PDF).
>   OK, next question:  is there an easier way than the following
> to create a 3D perspective plot with reference grids on some faces?
> I  can pull the 3D grid code out of the rgl or scatterplot3d
> packages and reimplement it here, I guess ...  another way to
> hack this might (?) be to play with tick lengths?

(I had been meaning to reply, but it slipped my mind.)

I can't see an easy way to do this for persp, but it's technically
possible (though not a one-liner) for cloud and wireframe. I can look
into it if you think it will help.

-Deepayan