Re: [R] Question about is.na for missing data

2009-10-29 Thread Karl Ove Hufthammer
On Thu, 29 Oct 2009 18:12:34 +0100 Karl Ove Hufthammer wrote: > > rhs is the wrong length for indexing by a logical matrix > > > > An inspection of the data set reveals no value of the sort give above. > > Any thoughts? > > I would venture to guess that your 'asd' object is not of the type

Re: [R] Question about is.na for missing data

2009-10-29 Thread Karl Ove Hufthammer
On Thu, 29 Oct 2009 11:45:19 -0500 David Kaplan wrote: > asd[is.na(asd)] <- mean(asd,na.rm=TRUE) > > This has worked quite well in the past. Now, however, I am getting the > following error messsage. > > Error in `[<-.data.frame`(`*tmp*`, is.na(asd), value = > c(2433.64150943396, : > rhs

[R] Question about is.na for missing data

2009-10-29 Thread David Kaplan
Greetings, I am trying to a simple mean imputation using the following command (yes, I know it is not an ideal way to impute missing data) asd[is.na(asd)] <- mean(asd,na.rm=TRUE) This has worked quite well in the past. Now, however, I am getting the following error messsage. Error in `[<-