Message-ID: <Pine.GSO.4.21.0003171536410.29432-100000@mail.biostat.washington.edu>
Date: 2000-03-17T23:38:48Z
From: Thomas Lumley
Subject: d-p-q-r-tests: why is plnorm(exp(Inf)) equal pnorm(Inf)?
In-Reply-To: <38d2b9ec.thoffman.zappa@zappa.sax.de>
On Sat, 18 Mar 2000, Thomas Hoffmann wrote:
> I stumbled across a result of d-p-q-r-tests:
>
> d-p-q-r-tests expects pnorm(Inf) and pnorm(-Inf) to give the same results as plnorm(exp(Inf))
> and plnorm(exp(-Inf)), respectively.
>
> Does this mean that all expressions result in NaN (which I expect for
> the 3rd and 4th expression only) or does it mean that R can prevent
> the exp(Inf) to be evaluated before being parsed in the context of
> plnorm?
No, it means that exp(Inf) and exp(-Inf) are evaluated correctly (Inf and
0 respectively) and the lognormal cdf is then evaluated as 1 and 0
respectively.
If the result were NaN then equality would not hold. Nothing is equal to
NaN, not even itself.
-thomas
Thomas Lumley
Assistant Professor, Biostatistics
University of Washington, Seattle
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._