[R] names( d$columnname )

[Duncan Murdoch]

On 4/22/2007 5:46 PM, ivo welch wrote:
> dear R wizards --- would it make sense for names(d$columnname) to be
> "columnname"?  I can preserve the columnname through x=subset(dataset,
> select="columnname"), of course, but it would seem that x=d$columnname
>  could also do this.  No?  Sincerely,  /iaw

If I understand what you're asking, I don't think so.

d$columnname

extracts an element of the list d.  It's typically some sort of vector, 
and vectors don't know what their own name is.

names(d$columnname)

looks for a names attribute on that vector.  If it doesn't have one, 
you'll get NULL.

There are plenty of functions in R which depend on the form of the 
argument passed to them, not just its value, but we shouldn't add more 
without a very good reason.  The change you ask for would mess up the 
following calculation:

 > x <- 1:26
 > names(x) <- letters
 > d <- list(columnname=x)
 > names(d$columnname)
  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o" "p" 
"q" "r"
[19] "s" "t" "u" "v" "w" "x" "y" "z"


Duncan Murdoch