On Nov 17, 2008, at 22:48 , Steve Lianoglou wrote:
On Nov 17, 2008, at 9:04 PM, Bill Northcott wrote:
On 18/11/2008, at 12:37 PM, Steve Lianoglou wrote:
I'm curious to see if using StatET is useful, too ... as this is
quite eclipse/plugin specific, perhaps you'd get better feedback
from the StatET mailing list?
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/statet-user
I started there but they have no clue about MacOS.
Oh ... I didn't see any messages related to this in their archives,
so I thought maybe you didn't find them, sorry.
I was looking for someone who might understand how this stuff
normally works on a Mac and hence I might be able to see what is
going wrong.
When you create a run config for R (Run > Run Confgurations ...), in
the "Main" tab, you have a drop down box with "Launch" options. The
Rterm option works, the RMI/JRI option doesn't.
Ideally I guess we want RMI/JRI ... here's what happens.
Launch R w/ Rterm
===============
I added --no-save in the Options/Arguments box, but I'm not sure
that matters, anyway, when you run the config, this works:
R> > Sys.getenv('R_SESSION_TMPDIR')
Sys.getenv('R_SESSION_TMPDIR')
R_SESSION_TMPDIR
"/var/folders/kL/kLbJmLNyEWa6w5gv6nb60U+++TI/-Tmp-//Rtmpp01O7b"
R> help.start()
help.start()
Making links in per-session dir ... done
If '/usr/bin/open' is already running, it is *not* restarted, and you
must switch to its window.
Otherwise, be patient ...
# Browser pops up with the page you want
Launch R w/ RMI/JRI
=================
Yes: bizarre. I can't get anything to work.
I can effectively get the R_SESSION_TMPDIR to be "what I want" by
setting the TMPDIR environment variable in the Run config's
Environment tab. For instance, I have TMPDIR set to "/Users/stavros/
tmp". The R_SESSION_TMPDIR var is then:
R> Sys.getenv('R_SESSION_TMPDIR')
R_SESSION_TMPDIR
"/Users/stavros/tmp/RtmpS1zNun"
I'm still getting some the same errors, though, as the R process
isn't making the required directories. I wonder if it has something
to do w/ java communicating the R process via RMI ... I don't really
know.
It is attempting to access this .R directory in the location
defined by $R_SESSION_TMPDIR. At what point might R be expected
to create that? When I run R in a Terminal these directories get
created.
So, after playing w/ this for a bit, I guess that:
(1) The $R_SESSION_TMPDIR is just created by monkeying with the
environments defined TMPDIR variable, which we can change in the Run
config's settings (like above)
(2) The "RtmpS1zNun" is created when R starts up
(3) The .R directory is then created after help.start() is executed.
I got (2) and (3) when running from the Terminal.
You know: I think I've realized what's going on, and I'm not sure
how to fix it. It looks as if the running R process doesn't have the
appropriate R environment variables set. For instance, in the RMI/
JRI mode, look:
R> Sys.getenv('USER')
USER
""
Likewise, the "help-links.sh" (in "/Library/Frameworks/R.framework/
Versions/2.8/Resources/share/sh") script that the help.start()
function uses requires environment variables like "R_DOC_DIR",
"USER_R_HOME", to be set. All of these are empty. I tried setting
the "R_DOC_DIR" to the appropriate location in the Run config's
environment tab, but that's still not enough.
Apparently the way Eclipse is starting R is wrong. The recommended way
is to use the R shell wrapper even with JRI -- usually "R CMD foo"
will do the trick where foo is the script that starts Java.
Cheers,
S
I can't really spend much more time on this now, but I think the
problem is somewhere buried in there ...
-steve
--
Steve Lianoglou
Graduate Student: Physiology, Biophysics and Systems Biology
Weill Medical College of Cornell University
http://cbio.mskcc.org/~lianos
_______________________________________________
R-SIG-Mac mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-mac
_______________________________________________
R-SIG-Mac mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-mac
