Skip to content

no warning with NULL component in data.frame

2 messages · Emmanuel Paradis, Thomas Lumley

Emmanuel Paradis
# 
Dear all,

I sent a message a few days ago about NULL components in list or
data.frame. To reiterate, if I have a list or a data.frame, and try to
manipulate some elements using the $ symbol, no warning is given if the
elements do not exist.

	> DF <- data.frame(x=c(1,2), y=c(3,4))
	> DF
	  x y
	1 1 3
	2 2 4
	> tmp <- c(DF$u, DF$v)
	> tmp
	NULL

Practically, I met the problem recently reading some data with:

	DF <- read.table(..., header=TRUE)

where the column names were L1, L3, L4, L5, L6 (L2 missing), and then went
on concatenating all these:

	X <- c(DF$L1, DF$L2, DF$L3, DF$L4, DF$L5, DF$L6)

which, of course, gave no warning, but then I realized that the length of X
was not what I expected.

A work-around to this particular problem would be to attach the data.frame:

	> attach(DF)
	> tmp <- c(u, v)
	Error: Object "u" not found

But yet, I do not understand the reason for this:

	> DF$anything.you.could.imagine
	NULL

since I would expect something like:

	> DF$anything.you.could.imagine
	Error: Component "anything.you.could.imagine" not found in Object "DF"

I suppose this has more to do with the design of the language. I understand
too that components within lists or data frames are not objects. 

Best,

Emmanuel Paradis
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Thomas Lumley
#
On Wed, 31 May 2000, Emmanuel Paradis wrote:

            

      
      
      
    

        
"It's not a bug, it's a feature"

You can do is.null(list$component) to test very efficiently whether the
component is there, which would be hard if it gave an error. Also, it
turns out to be useful to be able to handle a list which may have missing
components without a lot of special cases and without errors.

Anyway, it's a fairly fundamental part of the language.

There might be a case for warning if you access a non-existent element of
a data frame. The problem here is that a lot of correct and functional 
code probably does access non-existent elements and the warnings would
get annoying.

Accessing non-existent dataframe columns with "[" does give an error, so
you could do that instead.

            

      
      
      
    

        
NULL

            

      
      
      
    

        
Error in [.data.frame(trees, , "aardvark") : 
        subscript out of bounds


	-thomas

Thomas Lumley
Assistant Professor, Biostatistics
University of Washington, Seattle

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._