Re: [R] Problem about zero

From: Gabor Grothendieck <ggrothendieck_at_gmail.com>
Date: Tue, 15 Jun 2010 18:40:14 -0400

On Tue, Jun 15, 2010 at 6:00 PM, Yen Lee <b88207001_at_ntu.edu.tw> wrote:
> Hello, everyone,
>
> There's a problem about zero in R and I really need your help.
>
>
>
> I have a vector shown as x=c(0.1819711,0.4811463,0.1935151,0.1433675),
>
> The sum of this vector is shown as 1 in R, but when I type 1-sum(x), the
> value is not zero, but -2.220446e-16.
>
> I can accept that this value is quite small and could be seen as zero, but
> there would be a problem when it's not really zero but a negative value in
> my algorithm.
>
>
>
> Therefore I would like to know that how could it be avoid.
>
> One way I think is to define the value 1-sum(x) as zero when it is smaller
> than a particular value, but the particular value is not be set yet.
>
> I would like to know more about the definition of the shown zero in R.
>

As others have pointed out please read the R FAQ.

Actually on my machine: Windows Vista running C2D BLAS, "R version 2.11.1 Patched (2010-05-31 r52167)"
I do get zero:

> x <- c(0.1819711,0.4811463,0.1935151,0.1433675)
> 1-sum(x)
[1] 0

Also check out sum.exact in the caTools package.



R-help_at_r-project.org 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. Received on Tue 15 Jun 2010 - 22:43:10 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Tue 15 Jun 2010 - 23:00:33 GMT.

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

list of date sections of archive