atan2(1,1i)

[Brian Ripley]

BTW, this *really* is an R-devel topic, so please try to use the
appropriate list.

OK, I'm moving this matter to R-devel.

[Robin Hankin]

My point was that the documentation is inconsistent with the
function.

I think it is _your reading_ of the documentation that is inconsistent.
An equally plausible reading is that both x and y must be numeric or 
complex.

If a documentation may be read with many plausible but different 
meanings, then it is ambiguous and needs to be improved.  There is no 
point in asserting, as you did above, that the problem lies in the 
reading of the original poster.

The R documentation is quite good and very useful, there is a tremendous 
job in there, which I acknowlege.  Much thanks to all.  This being said, 
however, the documentation is not perfect, and you should stop reacting 
to R criticism as if it was personal offense, because it is not.

Fran?ois Pinard   http://pinard.progiciels-bpi.ca