[Rd] Issue when calling deparse(substitute(x)) from C with "anonymous" R vectors ?

From: Laurent Gautier <lgautier_at_gmail.com>
Date: Thu, 19 Nov 2009 08:40:12 +0100


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:

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_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel Received on Thu 19 Nov 2009 - 07:42:41 GMT

This archive was generated by hypermail 2.2.0 : Thu 19 Nov 2009 - 11:50:32 GMT