[R] Assigning NA to a rows of a dataframe/datamatrix

[David Winsemius]

On May 25, 2010, at 5:23 AM, john james wrote:

> Dear R-users,  I have a problem, I have the following dataframe:
>
> d<-data.frame(
>  'y1'=c(1,2,1,2,1,NA,NA),
> 'y2'=c(1,2,1,1,1,2,1),
> 'y3'=c(1,NA,1,NA,NA,2,1),
> 'y4'=c(NA,2,NA,1,1,2,NA),
> 'a'=c(1,1,1,1,1,1,2)
> )
>
> where the last variable counts the number of missing values in a  
> row. Now, i want to set rows where a>1 to NA and arrive at something  
> like the following;
>
> dnew<-data.frame(
>  'y1'=c(1,2,1,2,1,NA,NA),
> 'y2'=c(1,2,1,1,1,2,NA),
> 'y3'=c(1,NA,1,NA,NA,2,NA),
> 'y4'=c(NA,2,NA,1,1,2,NA),
> 'a'=c(1,1,1,1,1,1,4)
> )
>
> Please, how do I go about this. Many thanks!

is.na(d[rowSums(is.na(d))>1, 1:4]) <- TRUE

If you want the column "a" to be updated you would need to recalculate  
the numbers of NA's

d[ , "a"] <- rowSums(is.na(d))

-- 
David

David Winsemius, MD
West Hartford, CT