Question

Given a list (a) and a numeric vector (b) of the same length as a:

a <- list(1, c(3, 7), 5)
b <- c(1,1,2)

How to create a new list of length max(b) where every element in a is put in the corresponding index according to b and combined if multiple occurrences occur, so that you get:

[[1]]
[1] 1 3 7

[[2]]
[1] 5

This should hopefully be done very efficiently (no loops), as the dimensions of my data tends to be very big…

I’m quite sure there is an obvious solution I haven’t thought about…

Thanks for the help

Answer 1

What about:

split(unlist(a), rep(b, sapply(a, length)))
$`1`
[1] 1 3 7

$`2`
[1] 5