[Rd] format scientific + plotmath potential bug

Duncan Murdoch murdoch.duncan at gmail.com
Tue Jan 18 14:24:48 CET 2011


On 17/01/2011 5:06 PM, Philip Johnson wrote:
> I have run into a potential bug somewhere between format (specifically
> scientific notation) and plotmath that results in displaying:
> 	$1e+01^{2e+00}$
> instead of
> 	$10^2$
>
> Reproduce by:
> 	plot.new()
> 	a=format(10, scientific=TRUE)
> 	mtext(expression(10^2), line=1) # looks like $1e+01^{2e+00}$
> 	10 # this fixes the problem on the next line
> 	mtext(expression(10^2), line=2) # looks like $10^2$
>
> I can narrow the trigger somewhat further by replacing the "a=..." line
> with:
> 	a=.Internal(format(10, FALSE, NULL, 0L, NULL, 3L, TRUE, TRUE))
> Tracing this call into the C started giving me a headache, so I'm hoping
> that one of the R core gurus can confirm&  file a bug report if necessary.
>
> I ran into this on Ubuntu / Lucid using R version 2.10.1 (2009-12-14).
> I confirmed it still exists in R 2.12.1 (2010-12-16) on a fresh install
> of Ubuntu / Maverick.

Yes, I see this too in R-devel and 2.12.1.  Definitely a bug, and I've 
submitted it to the bug system:

https://bugs.r-project.org/bugzilla3/show_bug.cgi?id=14477

If you want to be kept up to date about fixes you can add yourself to 
the CC list there.

Duncan Murdoch



More information about the R-devel mailing list