On Wed, 25 Jul 2007, ONKELINX, Thierry wrote: >> -----Oorspronkelijk bericht----- >> Van: Prof Brian Ripley [mailto:[EMAIL PROTECTED] >> Verzonden: woensdag 25 juli 2007 12:20 >> Aan: ONKELINX, Thierry >> CC: [email protected] >> Onderwerp: Re: [R] Strange warning in summary.lm >> >> On Wed, 25 Jul 2007, ONKELINX, Thierry wrote: >> >>> Dear Peter, Uwe and Brian, >>> >>> I've found some more problems with options(OutDec = ","). >>> >>> 1) as.numeric yields NA where it shouldn't >> >> It should: where does it say otherwise? OutDec affects output, only. > > I was doing something like > >> options(OutDec = ",") >> df <- data.frame(var = rep(1:8/4, 10), x = rnorm(80)) >> df.a <- aggregate(df$x, by = list(var = df$var), FUN = sum) >> as.numeric(df.a$var) > [1] NA NA NA 1 NA NA NA 2 > Warning message: > NAs introduced by coercion in: as.double.default(df.a$var) > > because I needed df.a$var as numeric again.
And as Martin maechler has pointed out, we don't (and can't because ',' is used for other things) support that. > >>> 2) anova yields the same warning as summary >> >> It does not in current R: try any fairly recent version of R-devel. >> >> Please don't keep reporting a problem we have already fixed, >> as the FAQ and the posting guide explicitly ask of you. > > Sorry about that. I wasn't aware that the error was in a fixed > subroutine that is used by summary and anova. > I'm hestating to use a R-devel version because I haven't tried > installing R from a tar.gz (I'm one of those lazy windows users that > prefer the .exe version). For now I think I'll return to the good old > option(OutDec = "."). Binary versions are available on CRAN. -- Brian D. Ripley, [EMAIL PROTECTED] 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 ______________________________________________ [email protected] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
