[Rd] cbind(deparse.level=2,...) problems

[Peter Dalgaard]

William Dunlap wrote:
> Should the deparse.level=2 argument to cbind
> and rbind be abandoned?  It is minimally documented,
> not used in any CRAN package, and causes some problems.

Bill,

This code has been touched by Brian quite recently. Try again with a 
current r-devel checkout. (Specifically, later than r51007).

-p

> E.g.,
> 
> (a) If a matrix input has row names but not column names
>     cbind(deparse.level=2,...) stops.
> m<-matrix(11:14,nrow=2,ncol=2,dimnames=list(Row=c("R1","R2"),Col=charact
> er()))
>> cbind(m, 101:102, deparse.level=2)
> Error in cbind(m, 101:102, deparse.level = 2) :
>   SET_STRING_ELT() can only be applied to a 'character vector', not a
> 'NULL'
> 
> (b) If no argument is tagged then deparse.level=2
>     makes no column names, while deparse.level=1 makes
>     names for for the simple cases (where the argument
>     is a name):
>> x<-1:3
>> cbind(x, 11:13, deparse.level=1)
>      x
> [1,] 1 11
> [2,] 2 12
> [3,] 3 13
>> cbind(x, 11:13, deparse.level=2)
>      [,1] [,2]
> [1,]    1   11
> [2,]    2   12
> [3,]    3   13
> 
>      (If one argument is tagged, teens=11:13, then
>      deparse.level=2 will name columns that
>      deparse.level=1 will not, like x+1, which I
>      think is the intended behavior.)

-- 
    O__  ---- Peter Dalgaard             Øster Farimagsgade 5, Entr.B
   c/ /'_ --- Dept. of Biostatistics     PO Box 2099, 1014 Cph. K
  (*) \(*) -- University of Copenhagen   Denmark      Ph:  (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)              FAX: (+45) 35327907