I also see this. It seems that the 100th point plotted is not connected to the
101st, regardless of the actual data values. A more visually dramatic example is
x<-seq(0,8*pi,length=102)
plot(x,cos(x),type="l")
which makes it clear that the problem isn't just something like antialiasing.
On further research it looks as though this has already been fixed (though it
wasn't mentioned in NEWS, so it wasn't immediately obvious).
You could try the patched build from http://r.research.att.com/
-thomas
Thomas Lumley Assoc. Professor, Biostatistics
[EMAIL PROTE
On Sun, 10 Aug 2008, Thomas Lumley wrote:
On further research it looks as though this has already been fixed (though it
wasn't mentioned in NEWS, so it wasn't immediately obvious).
You could try the patched build from http://r.research.att.com/
I don't see it on my R-patched build. R-sig-
