Let's not speculate when doing basic research is so easy. I guess the
thread about 64-bit Windows referred to was
https://stat.ethz.ch/pipermail/r-devel/2010-January/056301.html
https://stat.ethz.ch/pipermail/r-devel/2010-January/056411.html
(note the update).
The CRAN daily check
http://cran.r-project.org/web/checks/check_summary.html shows that
106/2170 packages are currently giving errors, and 20 or so of those
are available from CRANextras. See
http://cran.r-project.org/bin/windows64/contrib/2.11/@ReadMe for more
details.
Note that the x64 Windows builds are for the unreleased R pre-2.11.0:
they will become more visible on CRAN once that is released.
For MacOS X: the CRAN build of 2.10.x for 10.5/6 includes the 64-bit
x86_64 architecture, and many packages are available (about the same
proportion as on x64 Windows).
For Unix-alikes (including Linux), 64-bit R has been available for
about a decade and some of us have been running 64-bit R exclusively
for years. Again, the CRAN Daily Check page shows a high rate of
success on x86_64 Linux.
Installation of packages without compiled code is done entirely in R
and so is the same on any platform. The difficulties in installing
packages arise either with compiled code or with dependencies: for
example as we do not yet have rgdal working on x64 Windows, none of
the dozen or so packages which depend on it are available.
On Sun, 21 Feb 2010, Sharpie wrote:
Axel Urbiz wrote:
Dear R users,
I know this issue came up in the list several times. I?m currently
running
R on 32-bit on Windows and due to memory limitation problems would like to
move to a 64-bit environment. I?m exploring my options and would
appreciate
your expertise:
1) Windows 64-bit: Prof. Brian Ripley recently posted the
experimental
built of R for win 64-bit. I?ll appreciate any feedback on anyone who has
been testing this. He also mentioned that for now, ?...this as only being
of
interest for those who only use a few relatively simple packages?. But if
one uses packages beyond those ?relatively simple?, how possible is today
to
have those installed?
As far as I know, the problems with running 64-bit R on windows are due to
the lack of a well-developed, 64 bit minGW compiler. Therefore I would
imagine that the packages that may have trouble installing under the
experimental 64 bit build are those that include either C or Fortran source
code that needs to be compiled. You may also find that you need to compile
packages yourself-- prebuilt 64 bit versions for windows may not be
available on CRAN.
Axel Urbiz wrote:
2) MacOS or Unix. Sorry for my ignorance on this?but if I use any of
these environments on 64-bit and installed R on any of those, this is all
I
need to have R working on 64-bit. How about installing specialized
packages?
Are the packages on the CRAN repositories ?ready tho use? on these systems
or do I have to do any additional work to get them going?
Thanks in advance for your help!
Axel.
MacOS/Unix/Linux shouldn't have a problem with running 64 bit R and building
64 bit R packages-- this is because the GNU project provides a very mature
set of 64 bit compilers that these systems can use. I personally use 64 bit
R on Mac OS and have had no problems developing and installing packages for
the 64 bit system.
-Charlie
--
View this message in context:
http://n4.nabble.com/R-on-64-Bit-tp1563895p1564056.html
Sent from the R help mailing list archive at Nabble.com.
______________________________________________
R-help@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.
--
Brian D. Ripley, rip...@stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UK Fax: +44 1865 272595
______________________________________________
R-help@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.
