[R] performance gap between R 1.7.1 and 1.8.0

[Peter Dalgaard]

"Liaw, Andy" <andy_liaw at merck.com> writes:

> > From: Peter Dalgaard [mailto:p.dalgaard at biostat.ku.dk] 
> > "Liaw, Andy" <andy_liaw at merck.com> writes:
> > 
> > > I'm really out of ideas at this point.  Can anyone see the 
> > problem given the
> > > above, or at least tell me what to try next?
> > 
> > Hmm. Looking into the actual Rprof.out file is sometimes revealing.
> > Could you provide a link to it (I assume it might be a bit large for
> > email).
> 
> The files aren't that big, but it certainly isn't pretty (not intended for
> human consumption, I suppose).
> 
> I don't have a place to post it, so I'm attaching it as .tgz.  Hopefully it
> at least makes it to some...

Everyone directly addressed, I'd expect (i.e. not r-help). As far as I
can see the main story is the pattern

... "loadNamespace" "try"...".requirePackage" "getClass"..."flexmix" ...

i.e lots of calls to getClass trigger an attempt to load the namespace
for the package

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907