Search code examples
rfor-loopconventions

Avoid Loop in Slicing Operation

I have the following code that I execute using a for loop. Is there a way to accomplish the same without a for loop?

first_list <- c(1,2,3, rep(1,5), rep(2,5), rep(3,5), rep(4,5))
print(first_list)
[1] 1  2  3  1  1  1  1  1  2  2  2  2  2
[1] 3  3  3  3  3  4  4  4  4  4

breaks <- c(rep(1,3), rep(5,4))
values <- vector()
i <- 1
prev <- 1

for (n in breaks){
    values[i] <- sum(first_list[prev:sum(breaks[1:i])])
    i <- i + 1
    prev <- prev + n
}

print(values)
[1] 1   2   3   5   10   15   20

The purpose of the loop is to take the first three elements of a list, then add to that list the sum of the next four sets of 5.

Solution

  • You can use tapply for grouped operation

    tapply(first_list, rep(1:length(breaks), breaks), sum)

    or, preferably, using data.table

    library(data.table)
data.table(first_list, id=rep(1:length(breaks), breaks))[, sum(first_list), id]$V1

    If you have to perform it on your data as in your original post 

    setDT(mydata)
mydata[, id:=rep(1:length(breaks), breaks),][, sum(Freq), by=id]