Inno - 8 months ago 42

R Question

I'm trying to subset some values with the specific difference in a vector. In the followed the vector, I wanna separate a vector into several ones with specific difference of 1.

For example,

A problem

`a <- c(1, 1.2, 1.6, 2, 2.2, 2.6, 3, 3.2, 3.6, 4, 4.2, 4.6, 5, 5.2, 5.6, 6, 7, 8, 9, 10)`

As a result,

`b <- c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)`

c <- c(1.2, 2.2, 3.2, 4.2, 5.2)

d <- c(1.6, 2.6, 3.6, 4.6, 5.6)

I tried to code a For loop, but I think it's not efficient and there is the better method for solving this problem.

Answer

An alternative recursive solution, for every recursion, extract a smallest value based vector and pass the remaining values for the next recursion:

```
my_split = function(vec, tol) {
if(length(vec) == 0) list()
else {
mod1 <- (vec - min(vec))%%1
# here we check both abs(mod1) and abs(mod1 - 1) since for example
# (4.6 - 3.6)%%1 == 1 due to the fact that 4.6 - 3.6 < 1
splits <- split(vec, abs(mod1) < tol | abs(mod1 - 1) < tol)
c(list(splits$`TRUE`), my_split(splits$`FALSE`, tol))
}
}
my_split(a, 0.001) # use a tolerance here to deal with the problem that floating number
# can not be exactly represented
# [[1]]
# [1] 1 2 3 4 5 6 7 8 9 10
# [[2]]
# [1] 1.2 2.2 3.2 4.2 5.2
# [[3]]
# [1] 1.6 2.6 3.6 4.6 5.6
```