[R-pkg-devel] how to make "r-release-osx-x86_64-mavericks" machine support rgl 3D plots

Dear Duncan,

On 8 Oct 2015, at 22:10, Duncan Murdoch wrote:

On 08/10/2015 4:41 PM, Dr Gregory Jefferis wrote:

[snip]

I have a package which imports rgl i.e. rgl is loaded on package
startup. A number of examples (not tests) use rgl features and I do
not
want these to be run on CRAN.

Is there a way to configure rgl.useNULL to handle this situation?

I don't think so, but I don't think you really want that.

options(rgl.useNULL = TRUE)

doesn't stop rgl from running, it just stops it from trying to display
anything.  So you could put code like

if (!interactive())
options(rgl.useNULL = TRUE)

into your .onLoad function for your package, and it wouldn't try to
open
a display on CRAN, or anywhere else that wasn't interactive.

I tried following this suggestion, but this has the unintended (for me)
consequence that it is no longer possible to get rgl snapshots e.g. in
knitr when rgl.useNULL=TRUE? From what I can see writeWebGL works fine
but snapshot3d does not and this is not quite what I was expecting
(though it seems fair enough in retrospect).

Is there a workaround to get snapshots out of a RGL null display?

If not, I think I am back to square 1 in terms of trying to figure out
how to suppress rgl *examples* (tests are easy) on CRAN.

With many thanks,

Greg.

sessionInfo()

R version 3.2.3 (2015-12-10)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X 10.9.5 (Mavericks)

locale:
[1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] nat.nblast_1.6.1 catmaid_0.6      httr_1.0.0       nat_1.8.1
rgl_0.95.1435    devtools_1.9.1

loaded via a namespace (and not attached):
   [1] Rcpp_0.12.2        formatR_1.2.1      plyr_1.8.3
nabor_0.4.6        tools_3.2.3
   [6] nat.utils_0.5.1    digest_0.6.8       jsonlite_0.9.19
memoise_0.2.1      filehash_2.3
[11] igraph_1.0.1       curl_0.9.4         yaml_2.1.13        spam_1.3-0
          stringr_1.0.0
[16] knitr_1.11         dendroextras_0.2.1 grid_3.2.3         R6_2.1.1
          rmarkdown_0.7
[21] magrittr_1.5       codetools_0.2-14   htmltools_0.2.6
rsconnect_0.3.79   stringi_1.0-1





--
Gregory Jefferis, PhD
Division of Neurobiology
MRC Laboratory of Molecular Biology
Francis Crick Avenue
Cambridge Biomedical Campus
Cambridge, CB2 OQH, UK

http://www2.mrc-lmb.cam.ac.uk/group-leaders/h-to-m/g-jefferis
http://jefferislab.org
http://flybrain.stanford.edu

Is there a workaround to get snapshots out of a RGL null display?

No.

Theoretically it's possible to use a virtual frame buffer in X11 to 
hold the image, but I've got no experience with that myself.

If not, I think I am back to square 1 in terms of trying to figure 
out
how to suppress rgl *examples* (tests are easy) on CRAN.

You can test for a null device using rgl.useNULL().  If that is true, 
don't try to produce a snapshot.

On 30 Dec 2015, at 21:11, Duncan Murdoch wrote:

Is there a workaround to get snapshots out of a RGL null display?

No.

OK. Thanks for confirming.

Theoretically it's possible to use a virtual frame buffer in X11 to
hold the image, but I've got no experience with that myself.

I have used xvfb in some headless environments (including travis).

If not, I think I am back to square 1 in terms of trying to figure
out
how to suppress rgl *examples* (tests are easy) on CRAN.

You can test for a null device using rgl.useNULL().  If that is true,
don't try to produce a snapshot.

I wasn't being clear there. I want to have rgl example code (e.g. plot3d
statements) in my documentation

* that's not enclosed in conditionals
* or \dontrun
* but that doesn't use a live rgl window during package check
(especially on CRAN)
* ... unless I do something to request this.

The code is not actually making snapshots.

I think what I may try is changing my .onLoad to look something like
this:

.onLoad <- function(libname, pkgname) {

    DEVTOOLS_CRAN=toupper(Sys.getenv("NOT_CRAN")) %in% c("FALSE","0",
"NO")

    # this is a heuristic to determine if we are running on CRAN
    CRAN=toupper(Sys.getenv("_R_CHECK_INSTALL_DEPENDS_")) %in% c("TRUE",
"YES", "1")

    if(!interactive() && is.null(getOption('rgl.useNULL')) &&
(DEVTOOLS_CRAN || CRAN)){
      options(rgl.useNULL=TRUE)
    }

}

Thanks again,

Greg.

--
Gregory Jefferis, PhD
Division of Neurobiology
MRC Laboratory of Molecular Biology
Francis Crick Avenue
Cambridge Biomedical Campus
Cambridge, CB2 OQH, UK

http://www2.mrc-lmb.cam.ac.uk/group-leaders/h-to-m/g-jefferis
http://jefferislab.org
http://flybrain.stanford.edu

On 30/12/2015 4:42 PM, Dr Gregory Jefferis wrote:

On 30 Dec 2015, at 21:11, Duncan Murdoch wrote:

I wasn't being clear there. I want to have rgl example code (e.g. 
plot3d
statements) in my documentation

* that's not enclosed in conditionals
* or \dontrun
* but that doesn't use a live rgl window during package check
(especially on CRAN)
* ... unless I do something to request this.

I'm not sure I understand why.  Why not write examples that produce 
rgl displays (or nothing with a null device)?  rgl itself has lots of 
these, and passes checks on most systems.  Currently it is supposed to 
detect display initialization failures and fall back to the null 
device, but I don't know if this works on the only CRAN system that 
seems to need it, the OSX system, because I haven't seen results of 
tests yet.

I think what I may try is changing my .onLoad to look something like
this:

I would be careful about that -- the CRAN people want to run realistic 
tests, and they will likely object if you are trying to act 
differently on their system than on your users' systems.