R help - Jun 2011 - Generic function to split a vector by user defined values

Hi all,

In follow up to my reply regarding splitting/grouping a vector:

  https://stat.ethz.ch/pipermail/r-help/2011-June/280361.html

it seems logical that a generic approach might be useful. So here is one
possibility, which I present for use and improvement as may be appropriate.

x : a vector
split: the value to use for the splits

splitVec <- function(x, split)
{
  is.na(x) <- x == split
  R1 <- rle(!is.na(x))
  split(x, rep(cumsum(R1$values) * R1$values, R1$lengths))[-1]
}


So with Dave's original integer vector:

x <- c(4, 5, 3, 0, 0, 0, 2, 4, 6, 4, 0, 0, 0, 2, 2, 0, 3, 4, 1, 0)
> splitVec(x, 0)
$`1`
[1] 4 5 3

$`2`
[1] 2 4 6 4

$`3`
[1] 2 2

$`4`
[1] 3 4 1


With a character vector:

set.seed(1)
Vec <- sample(c(letters[1:4], "Z"), 10, replace = TRUE)
> Vec
 [1] "b" "b" "c" "Z" "b"
"Z" "Z" "d" "d" "a"
> splitVec(Vec, "Z")
$`1`
[1] "b" "b" "c"

$`2`
[1] "b"

$`3`
[1] "d" "d" "a"


Hope that this might be useful to folks.

Regards,

Marc Schwartz