[R] referring to a row number and to a row condition, and to columns simultaneously

[Dimitri Liakhovitski]

Hello, dear R-ers!

I have a data frame:
x<-data.frame(a=c(4,2,4,1,3,4),b=c(1,3,4,1,5,0),c=c(NA,2,5,3,4,NA),d=rep(NA,6),e=rep(NA,6))
x

When x$a==1, I would like to replace NAs in columns d and e with 8 and
9, respectively
When x$a != 1, I would like to replace NAs in columns d and e 101 and
1022, respectively.

However, I only want to do it for rows 2:5 - while ignoring what's
happening in rows 1 and 6.

Here is what I've come up with:

x
for(i in 2:5){
  x[i & x[[1]]==1,4:5]<-c(8,9)
  x[i & x[[1]]!=1,4:5]<-c(101,102)
  }
x

However, something is wrong here.
First, rows 1 and 6 are not ignored.
Second, the order of 101 and 102 changes - I, however, always want to
see 101 in column d and 102 in column e.

Any advice?
Thanks a lot!

-- 
Dimitri Liakhovitski
Ninah.com
Dimitri.Liakhovitski at ninah.com