[Rd] RE: [R] Removing "row.names"

Kurt Hornik Kurt.Hornik@ci.tuwien.ac.at
Wed, 7 Feb 2001 18:20:29 +0100

>>>>> Thomas Lumley writes:

> On Wed, 7 Feb 2001, Kurt Hornik wrote:
>> Correct.  Let me rephrase the question: how can it be possible to set
>> the (variable) names in a data frame to something non-unique?

> 1) he didn't: there was only one column

Maybe I should be more careful when composing my emails :-)

I think I am also concerned about data frames with variable names which
are not syntactically valid, but one can also get that by creating the
data frame with `check.names = FALSE'.

> 2) we allow duplicated column names
>  For example
>   a<-data.frame(rnorm(10),rnorm(10)) 
>  gives a data frame with two columns having the same name. What do you
> think they should be called?  
> We discussed this some time ago in the context of offsets in glm(), and
> model.offset contains code to handle multiple "(offset)" columns.

Hmm, I must have missed that.

My understanding thus far was that I can always rely on unique case and
variable names in a data frame.  I see the potential usefulness in
allowing multiple "(offset)" columns in a model frame ... but wouldn't
that (at least in theory) call for using a different class?

r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch