Jeffrey Horner wrote:
[...]
> This raises an interesting topic for me:
>
> In R-2.3.1, .First.sys() autloads() a few functions from the various
> default packages (utils, stats, methods, graphics) IF those packages
> weren't specified in the default packages option (or the environment
> variable R_DEFAULT_PACKAGES). For instance, if a user doesn't want the
> utils package automatically loaded, then .First.sys() does this:
>
> sch <- search()
> if(! "package:utils" %in% sch) {
> autoload("example", "utils")
> autoload("methods", "utils")
> autoload("str", "utils")
> }
>
> which will automagically load utils if a call to example, methods, or
> str is made (why just those three? why not others in utils). I presume
> there was some heuristic to determine exactly which functions were
> autoloaded.
To answer this, all I had to do is look at the subversion log:
-------------------------------------------------------------------------
r27733 | ripley | 2003-12-21 04:21:33 -0600 (Sun, 21 Dec 2003) | 3 lines
allow all the standard examples to run with R_DEFAULT_PACKAGES=NULL
by judicious autoloading
-------------------------------------------------------------------------
That makes sense. But what about the change for R-trunk?
>
> Now in R-trunk today, I see that there are no autloaded functions set up
> from .First.sys() for any of the default packages. So if R is started
> with R_DEFAULT_PACKAGES=NULL, then no default packages will be loaded
> and no functions will be set up via autoload(). Why?
>
> Jeff
--
http://biostat.mc.vanderbilt.edu/JeffreyHorner
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
