shift elements later. Character or factor
variables are padded with
"", numerics with
NA.
Lag(x, shift = 1)
A.ttributes of the original object are carried along to the new lagged one, but factor vectors are converted to character.
x
Frank Harrell
Lag(1:5,2)
Lag(letters[1:4],2)
Lag(factor(letters[1:4]),2)
# Find which observations are the first for a given subject
id <- c('a','a','b','b','b','c')
id != Lag(id)
!duplicated(id)