R devel - Feb 2013 - How to use summary.mer inside a package?

I have a question regarding the build of my project papeR (hosted on 
R-forge http://r-forge.r-project.org/R/?group_id=1574) with respect to 
lme4.

Both, Windows and MacOS are complaining that lme4 doesn't export summary:

   Error : object 'summary' is not exported by 'namespace:lme4'
   ERROR: lazy loading failed for package 'papeR'

Linux however builds the project as desired.

I now did check my package using Uwe Ligges' winbuilder project and got 
positive results i.e. no errors for Windows there. See (for the next 72 
hours):

http://win-builder.r-project.org/y3uwwx93nJ92
http://win-builder.r-project.org/10mtpspKx7us

Can anyone tell me why this happens (only on R-forge and only for 
certain systems) and how to prevent this?

[I already asked the R-forge maintainers and was redirected to the 
general R mailing lists]

As a side note: The reason for importing lme4's summary is that 
otherwise my function summary.fixef, which extracts fixed effects from 
nlme and lme4 models for printing e.g. in reports, doesn't use/find 
summary.mer on "mer" objects -- even if lme4 is loaded earlier. On the
other hand, lme models (from package nlme) are handled correctly despite 
the fact that I do not import the corresponding summary function.

Thank you very much
   Benjamin

On 05.02.2013 12:01, Benjamin Hofner wrote:
> I have a question regarding the build of my project papeR (hosted on
> R-forge http://r-forge.r-project.org/R/?group_id=1574) with respect to
> lme4.
>
> Both, Windows and MacOS are complaining that lme4 doesn't export
summary:
>
>    Error : object 'summary' is not exported by
'namespace:lme4'
>    ERROR: lazy loading failed for package 'papeR'
>
> Linux however builds the project as desired.
>
> I now did check my package using Uwe Ligges' winbuilder project and got
> positive results i.e. no errors for Windows there. See (for the next 72
> hours):
>
> http://win-builder.r-project.org/y3uwwx93nJ92
> http://win-builder.r-project.org/10mtpspKx7us
>
> Can anyone tell me why this happens (only on R-forge and only for
> certain systems) and how to prevent this?
Perhaps the R-forge results are outdated? Looks like the R-forge builds 
(and hence probably also the checks) are stuck since some days.




> [I already asked the R-forge maintainers and was redirected to the
> general R mailing lists]
>
> As a side note: The reason for importing lme4's summary is that
> otherwise my function summary.fixef, which extracts fixed effects from
So, are summary.fixef and summary.mer methods or generics? If the 
former, why not call them via the generic?

Best,
Uwe Ligges



> nlme and lme4 models for printing e.g. in reports, doesn't use/find
> summary.mer on "mer" objects -- even if lme4 is loaded earlier.
On the
> other hand, lme models (from package nlme) are handled correctly despite
> the fact that I do not import the corresponding summary function.
>
> Thank you very much
>    Benjamin
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel