[R] plm: lag() and diff() do not (always) recognize a gap in the time dimension

[Christian Schoder]

Hello!

I came accross a strange behavior of the plm package. When using an
unbalanced panel with years lag() and diff() do not recognize a break in
the time dimension. It does, however, if there is only one more year
after the break. This is strange, right? 

Consider the following example:

> library(plm)
Loading required package: kinship
Loading required package: survival
Loading required package: splines
Loading required package: nlme
Loading required package: lattice
[1] "kinship is loaded"
Loading required package: Formula
Loading required package: MASS
Loading required package: sandwich
Loading required package: zoo

> dat.raw <- data.frame(id=c("b", "b","b", "c","c","c", "d","d","d","d",
"d","d","d"),
t=c(1980,1981,1983,1982,1983,1985,1984,1985,1988,1989,1993,1994,1995),
y=c(1,2,3,2,4,5,6,5,6,7,-2,1,3))

> dat.raw <- pdata.frame(dat.raw, index=c("id", "t"), drop.index=FALSE)

> dat.raw$l1<-lag(dat.raw$y,1)

> dat.raw$l2<-diff(dat.raw$y,1)

> dat.raw
       id    t  y l1 l2
b-1980  b 1980  1 NA NA
b-1981  b 1981  2  1  1
b-1983  b 1983  3 NA NA
c-1982  c 1982  2 NA NA
c-1983  c 1983  4  2  2
c-1985  c 1985  5 NA NA
d-1984  d 1984  6 NA NA
d-1985  d 1985  5  6 -1
d-1988  d 1988  6  5  1
d-1989  d 1989  7  6  1
d-1993  d 1993 -2  7 -9
d-1994  d 1994  1 -2  3
d-1995  d 1995  3  1  2

Why do the last two entries in the third and sixth rows indicate NA,
while row 11 does not seem to recognize the gap?

Would you have a solution to this problem for me? My dataset is huge and
I cannot edit it manually.

Thank you!

Christian