[R] choose() function returning anomalous results (zero instead of one)

[Petr Savicky]

On Thu, Sep 06, 2012 at 02:45:20PM -0700, Shayne Hodge wrote:
> I accidentally posted the wrong version of my code, omitting the Q1 loop,
> and, sure enough that loop generates non-integers:
> 
> qfracseq <- seq(from = 1, to = 0, by = -.1)
> Q <- 10
> Qseq <- Q*qfracseq
> 
> > Qseq [1] 10.00000000000000000  9.00000000000000000  8.00000000000000000  7.00000000000000000
>  [5]  6.00000000000000000  5.00000000000000000  3.99999999999999911
> 2.99999999999999911
>  [9]  1.99999999999999956  0.99999999999999978  0.00000000000000000
> 
> 
> -----
> 
> I'm a tad puzzled by why my code would do this; is this a result of some
> base-10 fractions not representing as cleanly in base-2 as in -10?

Hi.

Yes. For example, 0.2 in base 10 is 0.00110011001100110011001100110011...
in base 2.

See FAQ 7.31 and 

  http://rwiki.sciviews.org/doku.php?id=misc:r_accuracy

for further examples and some hints.

Hope this helps.

Petr Savicky.