Re: [Rd] Bug in rt() ? (PR#9422)

From: <ripley_at_stats.ox.ac.uk>
Date: Wed 20 Dec 2006 - 07:59:18 GMT


Why do you consider this might be a bug in R rather than in your expectations? Your 'expected' is not what others have expected ....

In 2.4.1 rt(ncp) operates continuously as ncp is varied through 0. That is different from saying that you want a central t (omitting ncp).

BTW, reports to R-bugs should *never* include a question mark: see the R FAQ. A question on R-devel (or even R-help) would have been reasonable.

On Tue, 19 Dec 2006, goodrich@fas.harvard.edu wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> <<insert bug report here>>
> Reproduced on Debian and Windows ...
>
> On 2.4.x if you execute
>
> set.seed(12345)
> t.1 <- rt(n = 1000, df = 20)
>
> set.seed(12345)
> t.2 <- rt(n = 1000, df = 20, ncp = 0)
>
> all.equal(t.1, t.2) ## Not close to true
>
> This appears to be due to the fact that in 2.4.x rt is now
>
> rt
> function (n, df, ncp = 0)
> {
> if (missing(ncp))
> .Internal(rt(n, df))
> else rnorm(n, ncp)/sqrt(rchisq(n, df)/df)
> }
> <environment: namespace:stats>
>
> Whereas in 2.3.1 rt() is verified to work as expected when someone
> (redundantly) types ncp = 0
>
> rt
> function (n, df, ncp = 0)
> {
> if (ncp == 0)
> .Internal(rt(n, df))
> else rnorm(n, ncp)/(rchisq(n, df)/sqrt(df))
> }
> <environment: namespace:stats>
>
> The only interface difference is missing(ncp) vs. ncp == 0 .
>
> - --please do not edit the information below--
>
> Version:
> platform = i486-pc-linux-gnu
> arch = i486
> os = linux-gnu
> system = i486, linux-gnu
> status =
> major = 2
> minor = 4.1
> year = 2006
> month = 12
> day = 18
> svn rev = 40228
> language = R
> version.string = R version 2.4.1 (2006-12-18)
>
> Locale:
> LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=en_US.UTF-8;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=en_US.UTF-8;LC_ADDRESS=en_US.UTF-8;LC_TELEPHONE=en_US.UTF-8;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=en_US.UTF-8
>
> Search Path:
> .GlobalEnv, package:car, package:rkward, package:stats,
> package:graphics, package:grDevices, package:utils, package:datasets,
> package:methods, Autoloads, package:base
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFFiF09zQDSXIcN85kRAhK4AKCBmPNkGszTzFdALfiGgW1LfNIOTwCfZ8RM
> 0gZjer8+rnmOPDFKZXBbFQw=
> =ff85
> -----END PGP SIGNATURE-----
>
> ______________________________________________
> R-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>

-- 
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

______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Received on Wed Dec 20 19:02:06 2006

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.1.8, at Wed 20 Dec 2006 - 08:31:04 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-devel. Please read the posting guide before posting to the list.