[Rd] Bug in xy.coords() or documentation error?

[Duncan Murdoch]

On 1/21/2006 1:21 PM, Gavin Simpson wrote:
> On Sat, 2006-01-21 at 13:12 -0500, Gabor Grothendieck wrote:
>> This was discussed just recently.   This is a design
>> error but the maintainers claim there are no cases of
>> interest where it matters.
> 
> Thanks Gabor,
> 
> I must have missed that discussion whilst I was on vacation. If what you
> say was the outcome of that discussion, it still means that the
> documentation for xy.coords is in error, as you may *not* provide a
> single argument 'x'.

You need to read the whole page.  You provide a single argument x by 
setting y to NULL, not by leaving it missing (as the docs said in 
earlier versions).

If you want to submit a patch to the docs, I'll take a look, but I don't 
like the one below.  It has basically the same flaw as the current docs: 
  you're providing two objects, not a single object.  Probably the 
clearest thing to do is just to say "see below" after the current 
wording, to point out that there are funny conventions here.

> If the intention is to keep the current behaviour - which is fine - then
> the documentation should be changed, perhaps along the lines of:
> 
>     x, y: the x and y coordinates of a set of points. Alternatively, a
>           single object 'x' can be provided if 'y = NULL' is also  
>           supplied.
> 
> 'object' might not be correct here - is a formula an 'object'?

Sure, why not?

Duncan Murdoch

> 
> Cheers,
> 
> Gav
> 
>> On 1/21/06, Gavin Simpson <gavin.simpson at ucl.ac.uk> wrote:
>>> Hi,
>>>
>>> I noticed the following problem with xy.coords() in R 2.2.1-patched
>>> (version info at the foot of this email) and R 2.3.0 unstable
>>> (subversion no: r37123):
>>>
>>>> xy.coords(x = matrix(1:20, ncol = 2))
>>> Error in xy.coords(x = matrix(1:20, ncol = 2)) :
>>>        argument "y" is missing, with no default
>>>> xy.coords(x = matrix(1:20, ncol = 2), y = NULL)
>>> $x
>>>  [1]  1  2  3  4  5  6  7  8  9 10
>>>
>>> $y
>>>  [1] 11 12 13 14 15 16 17 18 19 20
>>>
>>> $xlab
>>> [1] "[,1]"
>>>
>>> $ylab
>>> [1] "[,2]"
>>>
>>> And:
>>>
>>>> xy.coords(x = data.frame(x = 1:10, y = 1:10))
>>> Error in xy.coords(x = data.frame(x = 1:10, y = 1:10)) :
>>>        argument "y" is missing, with no default
>>>> xy.coords(x = data.frame(x = 1:10, y = 1:10), y = NULL)
>>> $x
>>>  [1]  1  2  3  4  5  6  7  8  9 10
>>>
>>> $y
>>>  [1]  1  2  3  4  5  6  7  8  9 10
>>>
>>> $xlab
>>> [1] "x"
>>>
>>> $ylab
>>> [1] "y"
>>>
>>> ... for example.
>>>
>>> ?xy.coords states:
>>>
>>>    x, y: the x and y coordinates of a set of points. Alternatively, a
>>>          single argument 'x' can be provided.
>>>
>>> Given that, I would have thought the above examples would have worked
>>> without explicitly passing y = NULL to xy.coords(). However, ?xy,coords
>>> later states:
>>>
>>>    If 'y' is 'NULL' and 'x' is a
>>>
>>> and the examples all illustrate the use of NULL passed as y.
>>>
>>> Is this a documentation error and a single argument x is not allowed, or
>>> is this a bug in the code? Either way, ?xy.coords contradicts itself as
>>> one would expect to be able to pass only x given the statement above.
>>>
>>> If this is a bug in the code, a potential workaround appears to be to
>>> change the first line of xy.coords from:
>>>
>>>   if (is.null(y)) {
>>>
>>> to
>>>
>>>   if (missing(y) || is.null(y)) {
>>>
>>> but I haven't tested this on anything other than my two examples.
>>>
>>> I will file a bug report if my observation is correct - but given all
>>> the erroneous bug reports lately, I thought I'd try my luck here where
>>> an error on my part would not cause the maintainers of the bug tracker
>>> any extra work.
>>>
>>>> version
>>>         _
>>> platform i686-pc-linux-gnu
>>> arch     i686
>>> os       linux-gnu
>>> system   i686, linux-gnu
>>> status   Patched
>>> major    2
>>> minor    2.1
>>> year     2006
>>> month    01
>>> day      18
>>> svn rev  37123
>>> language R
>>>> sessionInfo()
>>> R version 2.2.1, 2006-01-18, i686-pc-linux-gnu
>>>
>>> attached base packages:
>>> [1] "methods"   "stats"     "graphics"  "grDevices"
>>> [5] "utils"     "datasets"  "base"
>>>
>>> All the best,
>>>
>>> Gav
>>>
>>> --
>>> %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
>>> Gavin Simpson                     [T] +44 (0)20 7679 5522
>>> ENSIS Research Fellow             [F] +44 (0)20 7679 7565
>>> ENSIS Ltd. & ECRC                 [E] gavin.simpsonATNOSPAMucl.ac.uk
>>> UCL Department of Geography       [W] http://www.ucl.ac.uk/~ucfagls/cv/
>>> 26 Bedford Way                    [W] http://www.ucl.ac.uk/~ucfagls/
>>> London.  WC1H 0AP.
>>> %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
>>>
>>> ______________________________________________
>>> R-devel at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>>