I would like to create a vector in which each element is the i+6th element of another vector.
For example, in a vector of length 120 I want to create another vector of length 20 in which each element is value i, i+6, i+12, i+18... of the initial vector, i.e. I want to extract every 6th element of the original.
a <- 1:120 b <- a[seq(1, length(a), 6)] 
seq.int is faster in microbenchmarks, but I suspect that any performance increases in an actual program would be dwarfed by the running time of other parts.
a = 1:120; b = [::6]. Python can't do the former, R not the latter.Another trick for getting sequential pieces (beyond the seq solution already mentioned) is to use a short logical vector and use vector recycling:
foo[ c( rep(FALSE, 5), TRUE ) ] 
seq you have to be able to call length on the vector. letters[letters < 'm'][c(TRUE, FALSE, FALSE)]
I think you are asking two things which are not necessarily the same
I want to extract every 6th element of the original
You can do this by indexing a sequence:
foo <- 1:120 foo[1:20*6] I would like to create a vector in which each element is the i+6th element of another vector.
An easy way to do this is to supplement a logical factor with FALSEs until i+6:
foo <- 1:120 i <- 1 foo[1:(i+6)==(i+6)] [1] 7 14 21 28 35 42 49 56 63 70 77 84 91 98 105 112 119 i <- 10 foo[1:(i+6)==(i+6)] [1] 16 32 48 64 80 96 112 
foo[1:(i+6)!=(i+6)] i.e. output all values except the sixth.
x[1:(length(x)/n)*n]To select every nth element from any starting position in the vector
nth_element <- function(vector, starting_position, n) { vector[seq(starting_position, length(vector), n)] } # E.g. vec <- 1:12 nth_element(vec, 1, 3) # [1] 1 4 7 10 nth_element(vec, 2, 3) # [1] 2 5 8 11 To select every n-th element with an offset/shift of f=0,...,n-1, use
vec[mod(1:length(vec), n)==f] Of course, you can wrap this in a nice function:
nth_element <- function(vec, interval, offset=0){ vec[mod(1:length(vec), interval)==mod(offset, interval)] } I wanted to end up with a vector of the same length so I made this:
extract_nth_wraparound <- function(x, n) { index <- (seq_along(x) - 1) %% n + 1 ordered_indices <- order(index) return(x[ordered_indices]) } 
s = 6 # selection step, i.e., every s'th element will be selected p = 1 # starting position, i.e., position of the first element to be selected v = LETTERS # vector to select from v %>% .[seq_along(.) %% s == p] # [1] "A" "G" "M" "S" "Y" Check (position of the selected elements)
v %>% .[seq_along(.) %% s == p] %>% match(v) # [1] 1 7 13 19 25 From ?Arithmetic :
%% indicates x mod y (“x modulo y”), i.e., computes the ‘remainder’
