[R] Transferring a dataset from WINXP to Mac - problems with umlauts

Prof Brian Ripley ripley at stats.ox.ac.uk
Wed Mar 28 14:52:30 CEST 2012

On 28/03/2012 13:27, David Winsemius wrote:
> On a Mac running with a US locale, when I first type <option>-u (in
> Mail.app that shows a highlighted "naked umlaut) and then type 'u' I get:
> ü
> Similarly with opt-u, 'a'
> ä

I don't think that was the problem: he very likely has a German keyboard 
in Jena.  Here's my guess.

Your message dropped the contents for me, so pasting a bit from the 

'is it possible to correctly display umlauts in a dataset, which has 
been created under WINXP, on a Mac-system?'

Depends what is meant by 'a dataset'.  If this a a dataframe, I think 
the simplest thing to do would be to export it to csv (write.csv) on the 
Windows box, and then read it on the Mac with

read.csv(..., fileEncoding="cp1252")

(cp1252 being more plausible than latin1, and is a superset).

Or use dump() on the object, and then source(encoding = "cp1252")

It is also possible to save() the object on WinXP, load() on the Mac and 
then convert with iconv(), but that could be tedious if the non-ASCII 
characters are in more than one element.

If this is not the right guess, we need something more explicit ....

Brian D. Ripley,                  ripley at 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

More information about the R-help mailing list