[R-sig-ME] lme4 binary for OSX
Ben Bolker
bbolker at gmail.com
Wed Apr 10 01:47:43 CEST 2013
Federico Calboli <f.calboli at ...> writes:
>
> Dear Thierry,
>
> thank you for letting me know. I though this link
>
> http://lme4.r-forge.r-project.org/repos/bin/macosx/leopard/contrib/3.0/
>
> on the lme4 page meant it was, but that's not the case then.
>
> BW
>
> F
>
> On 9 Apr 2013, at 13:21, "ONKELINX, Thierry" <Thierry.ONKELINX at ...> wrote:
>
> > Dear Federico,
> > It seems that R-forge isn't yet updated to R 3.0.0. Hence the
> packages are not available for R 3.0.0
> > Best regards,
> >
> > Thierry
lme4 status report, R-3.0.0 related:
** STABLE version (lme4 on CRAN, lme4.0 on r-forge)
The MacOS binary build (only) on CRAN was missing because of some
failed (relatively obscure) tests -- problem still not diagnosed, but
for now I have commented these tests out. I've updated lme4.0 on
r-forge (it's still building, see status at
<http://r-forge.r-project.org/R/?group_id=60>) and have submitted the
updated version _as lme4_ to CRAN: if all goes well that should
propagate shortly (but not yet checked by the powers-that-be, so
there could still be some glitches to resolve)
I don't know if lme4.0-on-r-forge will be available for 3.0.0
or not once it successfully builds.
** DEVELOPMENT version (lme4 on github, and at lme4.r-forge repository)
BINARIES
I have added a recent MacOS binary at
http://lme4.r-forge.r-project.org/repos/bin/macosx/contrib/3.0/ .
This appears to be the correct path (install.packages(...) worked for
me on MacOS/R-3.0.0). I am still trying to sort the repository tree
out properly: the path above works, but
http://cran.r-project.org/doc/manuals/R-admin.html#
Setting-up-a-package-repository (broken URL!) suggests that MacOS
binaries should be "located at bin/macosx/snowleopard/contrib/3.y for
R versions 3.y.z" ... so I have added a copy there as well.
SOURCE
Still available on github for those with compile tools and devtools
package installed:
library("devtools")
install_github("lme4",user="lme4")
More information about the R-sig-mixed-models
mailing list