This is easy to reproduce in R:
chisq.test(c(1000, 1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008, 1009, 1010,
1011),
c(1000, 1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008, 1009, 1010, 1011))
The simple answer is: don't do that.
It is unclear what is a reasonable label to give in such a case: maybe
simply 'x' and 'y'?
On Thu, 19 Nov 2009, Laurent Gautier wrote:
Dear list,
When calling R from C, what appears like a spurious error can be triggered
during the execution of chisq.test(x, y).
This is happening when the following conditions are met:
- x and y are "anonymous" C-level R vectors (they do not have a symbol),
but they are protected from garbage collection
- x and y are "not too small" (it was experienced as soon as they are longer
than 17 elements).
The error is
Error in names(dimnames(x)) <- DNAME :
'names' attribute [4] must be the same length as the vector [2]
and can be traced to the use of deparse(substitute(x)) and
deparse(substitute(y)) in the R code.
What seems to be happening is that the deparse(substitute(x)) call
gives a character vector of length > 1 as soon as x is "not so small".
To demonstrate this, I am using rpy2 (as the problem was found when using it
http://sourceforge.net/mailarchive/forum.php?thread_name=4B043CA1.9050500%40salilab.org&forum_name=rpy-list
), but it will likely be present in other bridges to R as well.
#using R-2.10 and rpy2-2.1.dev
import rpy2.robjects as robjects
f = robjects.r('''function(x) return(deparse(substitute(x)))''')
tuple(f(robjects.FloatVector(range(17))))
('c(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)',)
# length 1
tuple(f(robjects.FloatVector(range(18))))
('c(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17', ')')
# length 2 !!
Does it seem to others like an issue present in other bridges as well ?
L.
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
--
Brian D. Ripley, rip...@stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UK Fax: +44 1865 272595
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
