Starting in SVN revision 47035 (which shows up in the R-2-9-0 line),
transform.data.frame() started accepting arguments like 'row.names' and
'stringsAsFactors' to be passed through to the data.frame() function.
It looks like this was an unintentional side-effect of letting multiple columns
be added properly.
Given that this has been implemented for quite a while, should it now be
documented? It's a little strange to support it though - the
'stringsAsFactors' argument might be handy, but 'check.names',
'row.names', and 'check.rows' are perhaps questionable. If
it's desirable to document the behavior, here's a possible patch.
---------------------------------------------
Index: src/library/base/man/transform.Rd
==================================================================---
src/library/base/man/transform.Rd (revision 61043)
+++ src/library/base/man/transform.Rd (working copy)
@@ -27,6 +27,10 @@
\code{_data}. The tags are matched against \code{names(_data)}, and for
those that match, the value replace the corresponding variable in
\code{_data}, and the others are appended to \code{_data}.
+
+ \code{transform.data.frame} also accepts the additional named
+ arguments that the \code{data.frame} function accepts,
+ e.g. \code{stringsAsFactors}.
}
\value{
The modified value of \code{_data}.
---------------------------------------------
--
Ken Williams, Senior Research Scientist
WindLogics
http://windlogics.com
________________________________
CONFIDENTIALITY NOTICE: This e-mail message is for the s...{{dropped:10}}
