[R] converting NA/non-NA's to a binary variable

[(Ted Harding)]

On 07-May-05 Gillian Rutherford wrote:
> Dear R colleagues,
> 
> I am trying to create a new column in a data frame, which
> converts values and NA's from another column into binary format.
> Essentially I need the NA's to become 1 and the rest to be 0.
> The code I wrote is returning the  following error message:
> 
> Error in if (mort[i, 4] != NA) mort[i, 8] <- 0 else if (mort[i, 4] == 
>:
>          missing value where TRUE/FALSE needed
> 
> The code is as follows:
> 
> for(i in 1:length(mort[,4]))
>       {
>               if(mort[i,4] != NA) mort[i,8] <- 0
>               else if(mort[i,4] == NA) mort[i,8] <- 1
>       }
> 
> I'd appreciate any advice or recommendations as to a better way of 
> achieving this.
> 
> Thanks
> Gillian

I think the following should do what you want, provided the column
mort[,8] exists:

  mort[,8] <- 0
  mort[is.na(mort[,4]),8] <- 1

Incidentally, tests like "== NA" or "!= NA" can produce unexpected
results! Use is.na() instead:

  tmp<-NA
  ## [1] NA
  tmp==NA
  ## [1] NA
  tmp!=NA
  ## [1] NA
  if(tmp==NA) 1 else 2
  ## Error in if (tmp == NA) 1 else 2 :
  ## missing value where TRUE/FALSE needed

  if(TRUE) 1 else 2
  ## [1] 1
  is.na(tmp)
  ## [1] TRUE

  if(is.na(tmp)) 1 else 2
  ## [1] 1
  if(!is.na(tmp)) 1 else 2
  ## [1] 2

Best wishes,
Ted.


--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.Harding at nessie.mcc.ac.uk>
Fax-to-email: +44 (0)870 094 0861
Date: 07-May-05                                       Time: 14:46:19
------------------------------ XFMail ------------------------------