[R] Odp: Data manipulation question

[Julien Barnier]

Hi Petr,

> d$z<-NA
> d$z[d$x %in% d$id] <- d$y[d$id %in% d$x]
>
> works in this particular case but it means you do not have multiple same 
> ids and X

Thanks for the idea. But the problem is that I can have multiple
ids...

In fact in the meantime I found a solution by using row names :

R> d
    id    x  y
1 0001 <NA> 21
2 0002 <NA> 13
3 0003 0001 45
4 0004 <NA> 71
5 0005 0003 20

R> rownames(d) <- d$id
R> d$z <- NA
R> d$z <- d[d$x,"y"]
R> d
       id    x  y  z
0001 0001 <NA> 21 NA
0002 0002 <NA> 13 NA
0003 0003 0001 45 21
0004 0004 <NA> 71 NA
0005 0005 0003 20 13


Thanks for your help,

Julien

-- 
Julien Barnier
Groupe de recherche sur la socialisation
ENS-LSH - Lyon, France