[R] failed to load MASS at start up

Prof Brian Ripley ripley at stats.ox.ac.uk
Sat May 3 09:42:39 CEST 2003


A belated follow-up on this -- I was on holiday over Easter.

On 16 Apr 2003, Douglas Bates wrote:

> Michael Na Li <lina at u.washington.edu> writes:
> 
> > Just installed R-1.7.0 (with recommended libraries) on RedHat 8.0.  
> > At R console, I can do
> > 
> > > library (MASS)
> > > 
> > 
> > just fine.  However, if I put a line 'library(MASS)' into ~/.Rprofile, it
> > fails to load,

Why would you want to do that?  There are options to set the list of
default packages, and

options(defaultPackages=c(getOption("defaultPackages"), "MASS"))

in .Rprofile is the preferred way to do this (and it works).

> > R : Copyright 2003, The R Development Core Team
> > Version 1.7.0  (2003-04-16)
> > ....
> > Type `q()' to quit R.
> > 
> > Error in get(x, envir, mode, inherits) : variable "biplot" was not found
> > Error in library(MASS) : package/namespace load failed
> > [Previously saved workspace restored]
> > 
> > >
> > 
> > What's happening?  I also tried to load 'lattice' (which I assume is where
> > biplot is defined) before 'MASS', but got the same error.
> 
> biplot is in mva

[...]

> By default mva is loaded but perhaps that does not occur until after
> ~/.Rprofile is executed.

That's true. Default packages are loaded right at the end of the startup 
options.

To fix this you could add the following line to .../library/MASS/NAMESPACE

importFrom(mva, biplot)

Unfortunately, that would load *two* copies of mva if MASS is loaded
before mva (as I understand it), which is why I removed it from the
NAMESPACE file.  I'll reinstate it for some future revision, possibly by 
a require(MASS) in a .onLoad.

-- 
Brian D. Ripley,                  ripley at 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



More information about the R-help mailing list