R devel - Feb 2003 - error from sample if sampling length one character (PR#2546)

Full_Name: Patrick Burns
Version: 1.6.1
OS: Suse Linux
Submission from: (NULL) (217.36.14.68)



Sample gives an error if it is sampling a length one character
vector:
> sample('a', 1)
Error in sample(x, size, replace, prob) : invalid first argument

The fix is trivial (numeric test in first line of the body):

function (x, size, replace = FALSE, prob = NULL)
{
    if (length(x) == 1 && is.numeric(x) && x >= 1) {
        if (missing(size))
            size <- x
        .Internal(sample(x, size, replace, prob))
    }
    else {
        if (missing(size))
            size <- length(x)
        x[.Internal(sample(length(x), size, replace, prob))]
    }
}
> sample.fixed('a', 1)
[1] "a"

> Sample gives an error if it is sampling a length one character
> vector:
>
> > sample('a', 1)
> Error in sample(x, size, replace, prob) : invalid first argument
>From the help page:
       x: Either a (numeric, complex, character or logical) vector of
          more than one element from which to choose, or a positive
          integer.

so this is as documented, hence not a bug.

[Re-created from the R-bugs page as my mail server is offline.]

-- 
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

Have you read the R FAQ on what a bug is?  It says

   Finally, a command's intended definition may not be best for
statistical analysis.  This is a very important sort of problem, but it is
also a matter of judgment.  Also, it is easy to come to such a conclusion
out of ignorance of some of the existing features.  It is probably best
not to complain about such a problem until you have checked the
documentation in the usual ways, feel confident that you understand it,
and know for certain that what you want is not available.

So, be honest: did you check the documentation?  If you did, why did not 
not mention that you knew it was a documented feature?

On Sun, 23 Feb 2003, Patrick Burns wrote:
> Just because it is documented does not mean it isn't a bug.
> 
> I found this in a real application that augments a character vector by 
> randomly
> selecting strings from a population.  In some circumstances that 
> population is a
> single string.
> 
> I don't see any reason not to make the change.
Let us enlighten you: programmers who read the documentation rely on the
current behaviour to catch user errors, as they are entitled to do.

I trust you have filed a bug report on the `real application' for
mis-using R.

-- 
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