R help - Nov 2000 - assigning to data frames with whole columns of NAs

I suppose this could be described as a feature (it seems to be similar in
S-Plus), but it looks to me more like a bug.  Why can't the assignment below
to a row of "emptyframe" (or "anotherframe") be made?  This
with R --vanilla (version info below).

Regards -- David
David Firth                        Phone +44 1865 278544
Nuffield College                   Fax   +44 1865 278621
Oxford OX1 1NF                     Secretary +44 1865 278612
United Kingdom                     Email david.firth at nuffield.oxford.ac.uk

> emptyframe<-data.frame(a=c(NA,NA),b=c(NA,NA))
> emptyframe
   a  b
1 NA NA
2 NA NA
> emptyframe[1,]<-c(1,2)
Warning messages: 
1: invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value =
vjj)
2: invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value =
vjj) 
> emptyframe
   a  b
1 NA NA
2 NA NA
> anotherframe<-data.frame(a=c(8,9),b=c(NA,NA))
> anotherframe
  a  b
1 8 NA
2 9 NA
> anotherframe[1,]<-c(1,2)
Warning message: 
invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value = vjj)
> anotherframe
  a  b
1 1 NA
2 9 NA
> yetanotherframe<-data.frame(a=c(8,NA),b=c(9,NA))
> yetanotherframe
   a  b
1  8  9
2 NA NA
> yetanotherframe[1,]<-c(1,2)   #  no problem with this one
> yetanotherframe
   a  b
1  1  2
2 NA NA
> version
         _                
platform i686-pc-linux-gnu
arch     i686             
os       linux-gnu        
system   i686, linux-gnu  
status                    
major    1                
minor    1.1              
year     2000             
month    August           
day      15               
language R 
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

David Firth <david.firth at nuffield.oxford.ac.uk> writes:
> > emptyframe<-data.frame(a=c(NA,NA),b=c(NA,NA))
> > emptyframe
>    a  b
> 1 NA NA
> 2 NA NA
> > emptyframe[1,]<-c(1,2)
> Warning messages: 
> 1: invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value
= vjj)
> 2: invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value
= vjj)
Problem, what problem?
> emptyframe  
   a  b
1 NA NA
2 NA NA
> emptyframe[1,]<-c(1,2)
> emptyframe
   a  b
1  1  2
2 NA NA

OK, I'm pulling your leg. The snag is that NA is logical NA, unless
something coerces it to something else, and when constructing data
frames, logical vectors are coerced to factors. So I had defined:
> emptyframe<-data.frame(a=as.numeric(c(NA,NA)),b=as.numeric(c(NA,NA)))
-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

David Firth wrote:
> 
> I suppose this could be described as a feature (it seems to be similar in
S-Plus), but it looks to me more like a bug.  Why can't the assignment below
to a row of "emptyframe" (or "anotherframe") be made?  This
with R --vanilla (version info below).
> 
> > emptyframe<-data.frame(a=c(NA,NA),b=c(NA,NA))
> > emptyframe
>    a  b
> 1 NA NA
> 2 NA NA
> > emptyframe[1,]<-c(1,2)
> Warning messages:
> 1: invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value
= vjj)
> 2: invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value
= vjj)
OK. That's not a bug! 
str(emptyframe) tells us: $a and $b are factors. You cannot assign the
*levels* "1" or "2" on this way.
So we can generate another similar example:
  emptyframe <- data.frame(a=c("A", "B"),
b=c("A", "B"))
  str(emptyframe)
  emptyframe[1,] <- c(1, 2) # Don't work either, what we expect.

> > emptyframe
>    a  b
> 1 NA NA
> 2 NA NA
> > anotherframe<-data.frame(a=c(8,9),b=c(NA,NA))
> > anotherframe
>   a  b
> 1 8 NA
> 2 9 NA
> > anotherframe[1,]<-c(1,2)
> Warning message:
> invalid factor level, NAs generated in: [<-.factor(*tmp*, iseq, value =
vjj)
> > anotherframe
>   a  b
> 1 1 NA
> 2 9 NA
> > yetanotherframe<-data.frame(a=c(8,NA),b=c(9,NA))
> > yetanotherframe
>    a  b
> 1  8  9
> 2 NA NA
> > yetanotherframe[1,]<-c(1,2)   #  no problem with this one
> > yetanotherframe
>    a  b
> 1  1  2
> 2 NA NA
Right, because here $a and $b are numeric: str(yetanotherframe)

If you really want to generate a data.frame only with NAs in it, you can
use for example:
  emptyframe <- data.frame(a=as.numeric(c(NA, NA)), b=as.numeric(c(NA,
NA)))
  emptyframe[1,] <- c(1, 2) # Works now ...

 
Uwe Ligges
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._