Thanks a lot for your prompt answer and your advice. Indeed, I managed
to track the bug by launching R from the c-shell instead of the Mac GUI.
In that case, after typing
I obtained the following error message:
> Erreur dans parse(file, n, text, prompt) : syntax error at
> 653: irr.name =3D "R
> Erreur : unable to load R code in package 'irr'
Because Mac, PC and Unix encode end-of-lines differently, this actually
corresponds to line # 1304 in the file "~/Library/R/library/irr/R/irr"
(it should be line # 653 if you work on a PC):
> rval <- list(method =3D "Mean of bivariate correlations R=FF",
> subjects =3D ns, raters =3D nr,
> irr.name =3D "R=FF", value =3D coeff)
The character "=FF" (displayed as a "?" in the c-shell) is clearly not=20=
I thus tried to replace it here by a normal "y", and similarly for=20
lines lower, and now the package loads without any problem.
I still need to understand how to avoid a conflict between the "icc"=20
defined both in the "irr" and "psy" packages, but this is another=20
feel free to give me any other advice!]
Thanks again; your sincerely,
Marina Saadia Otero
On May 6th, 2006, at 13:44, Prof Brian Ripley wrote:
> Please contact the package maintainer (Cc:ed here). =46rom the R FAQ:
> Finally, check carefully whether the bug is with R, or a contributed
> package. Bug reports on contributed packages should be sent first =
> the package maintainer, and only submitted to the R-bugs repository=20=
> package maintainers, mentioning the package in the subject line.
> and there is similar advice in the R posting guide.
> On Sat, 6 May 2006, [hidden email] wrote:
>> Full_Name: Marina Saadia Otero
>> Version: R 2.2.0
>> OS: Mac OS X 10.3.9
>> Submission from: (NULL) (220.127.116.11)
>> After having installed the "irr" package (from the
>> installation Menu of the Mac interface), the command
>> makes the R application quit. This has been tested from
>> a clean new session, and even from a brand new account.
>> Re-installing the package does not solve the problem.