Yes, it's a bug, but fortunately it's already fixed in the development
version of ggplot2. The problem arises from my incomplete knowledge
of floating point arithmetic - my previous check for zero range was
too imprecise.
Hadley
> --
> You received this message because you are subscribed to the ggplot2 mailing list.
> Please provide a reproducible example: http://gist.github.com/270442
>
> To post: email ggp...@googlegroups.com
> To unsubscribe: email ggplot2+u...@googlegroups.com
> More options: http://groups.google.com/group/ggplot2
>
--
Assistant Professor / Dobelman Family Junior Chair
Department of Statistics / Rice University
http://had.co.nz/