How to find the first smaller value compared to the current row in subsequent rows?

Question

Suppose this is the data:

data<-data.frame(number=c(4,5,3,1,0), 
             datetime=c(as.POSIXct("2015/06/12 12:10:25"),
                        as.POSIXct("2015/06/12 12:10:27"),
                        as.POSIXct("2015/06/12 12:10:32"),
                        as.POSIXct("2015/06/12 12:10:33"),
                        as.POSIXct("2015/06/12 12:10:35")))

  number   datetime
1      4 2015/06/12 12:10:25
2      5 2015/06/12 12:10:27
3      3 2015/06/12 12:10:32
4      1 2015/06/12 12:10:33
5      0 2015/06/12 12:10:35

I want to calculate the time between a row to the next smaller value. Desired output:

   number  next smaller   time between
1      4              3             7
2      5              3             5
3      3              1             1
4      1              0             2
5      0             NA          NA

Example: 3 is the first number in subsequent rows which is smaller than 4.

Any suggestion? package?

Answer 1

Well it's not pretty and probably not super efficient, but it seems to get the job done. Here we go ...

newcols <- with(data, {
    lapply(seq_along(number), function(i) {
        x <- number[-(1:i)][-i][1]
        c(x, abs(datetime[i] - datetime[number == x])[1])
    })
})

setNames(
    cbind(data[1], do.call(rbind, newcols)), 
    c(names(data)[1], "nextsmallest", "timediff")
)
#   number nextsmallest timediff
# 1      4            3        7
# 2      5            3        5
# 3      3            1        1
# 4      1            0        2
# 5      0           NA       NA