R help - Aug 2004 - Getting data loaded

Hi,

I have been informed of a bug in the concord package, in that the data files 
containing the tabulated critical values for Kendall's W are not loaded on 
the command library(concord).

I had assumed that the lines in install.R would correspond to the commands to 
load data in R, e.g.

data(Wcrit01)
data(Wcrit05)

While these work on the command line, I get the errors:

library(concord)
Warning messages:
1: Data set 'Wcrit01' not found in: data(Wcrit01)
2: Data set 'Wcrit05' not found in: data(Wcrit05)

The section on this in R-exts doesn't seem to have any information on how to
write the lines in install.R, nor does Checking and Building Packages. I 
managed to locate a few install.R files in other packages, but they were all 
empty. Any hints?

Jim

Dr Jim Lemon
Research Psychologist
Health Psychology Unit
University of Technology, Sydney

Feel free to ignore any garbage beneath this line.

-- 

DISCLAIMER: This email message and any accompanying attachme...{{dropped}}

I think you should load data in a .First.lib() function, or if 
you have a namespace, in a .onLoad() function.

-roger

Jim Lemon wrote:
> Hi,
> 
> I have been informed of a bug in the concord package, in that the data
files
> containing the tabulated critical values for Kendall's W are not loaded
on
> the command library(concord).
> 
> I had assumed that the lines in install.R would correspond to the commands
to
> load data in R, e.g.
> 
> data(Wcrit01)
> data(Wcrit05)
> 
> While these work on the command line, I get the errors:
> 
> library(concord)
> Warning messages:
> 1: Data set 'Wcrit01' not found in: data(Wcrit01)
> 2: Data set 'Wcrit05' not found in: data(Wcrit05)
> 
> The section on this in R-exts doesn't seem to have any information on
how to
> write the lines in install.R, nor does Checking and Building Packages. I 
> managed to locate a few install.R files in other packages, but they were
all
> empty. Any hints?
> 
> Jim
> 
> Dr Jim Lemon
> Research Psychologist
> Health Psychology Unit
> University of Technology, Sydney
> 
> Feel free to ignore any garbage beneath this line.
>

Diverted to R-devel: please see the posting guide.

On Wed, 18 Aug 2004, Roger D. Peng wrote:
> I think you should load data in a .First.lib() function, or if 
> you have a namespace, in a .onLoad() function.
Yes, indeed.  However, using data() for system data is not a good idea and 
for R 2.0.0 you can use the new sysdata.rda facility (documented in R-exts 
for that version-to-be).

The issue is that the code in install.R is run *before* the saved image in
the package is loaded (even though it is run after the package itself is
sourced), so you needed data(Wcrit01, package="concord").  When you
put
these in a .First.lib you should still specify the package for safety.

Finally, see the (currently draft) file

http://developer.r-project.org/200update.txt

> Jim Lemon wrote:
> > Hi,
> > 
> > I have been informed of a bug in the concord package, in that the data
files
> > containing the tabulated critical values for Kendall's W are not
loaded on
> > the command library(concord).
> > 
> > I had assumed that the lines in install.R would correspond to the
commands to
> > load data in R, e.g.
> > 
> > data(Wcrit01)
> > data(Wcrit05)
> > 
> > While these work on the command line, 
Don't think so.  You need to load your package first.
> I get the errors:
> > 
> > library(concord)
> > Warning messages:
> > 1: Data set 'Wcrit01' not found in: data(Wcrit01)
> > 2: Data set 'Wcrit05' not found in: data(Wcrit05)
> > 
> > The section on this in R-exts doesn't seem to have any information
on how to
> > write the lines in install.R, nor does Checking and Building Packages.
I
> > managed to locate a few install.R files in other packages, but they
were all
> > empty. Any hints?
Don't use install.R.
> > Jim
> > 
> > Dr Jim Lemon
> > Research Psychologist
> > Health Psychology Unit
> > University of Technology, Sydney
> > 
> > Feel free to ignore any garbage beneath this line.
> >
> 
> ______________________________________________
> R-help@stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html
> 
> 
-- 
Brian D. Ripley,                  ripley@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