uncover an integer sequence where all but one values are masked by NA's

Question

Say I have an integer vector where all but one entries are masked by NA's,

x <- c(NA, NA, 5, NA)

Is there a way to unmask those NA's (by in-place modification of NA or creating a new vector)? My desired output is

c(3, 4, 5, 6)

Answer 1

Could be

id <- which(!is.na(x))
newx <- seq_along(x) + (x[id] - id)

This works for your all NA's but one value vector.