[R] NAs introduced by coercion

[David Winsemius]

Wim Bertels <wim.bertels at khleuven.be> wrote in
news:1205233503.5326.26.camel at localhost: 

> Hallo,
> 
> i get a warning message that NAs are introduced by coercion,
> so my idea is to write a function to see which values are turned
> into NA 
> 
> For this i need to write a function to go through (loop) the
> original data and the transformed (with the introduced na) to see
> which data were transformed to NA.
> So the return of this function should be a 2*many matrix like
> structure, eg
> names: indexolddata, valueolddata
> 45, 789
> 89, 4568
> and so on
> (on the data that were transformed into an NA)
> 
> I am doing something wrong.

Try to build on this:

x <- data.frame(var1=1:10,var2=11:20)
x[5,2]<-NA
x[7,2]<-NA
row(x)[is.na(x)]

returns [1] 5 7

If you wanted the whole row, you would use that vector as a row index:

 x[row(x)[is.na(x)],]

returns:
  var1 var2
5    5   NA
7    7   NA

-- 
David Winsemius