Nearest neighbor in R

Question

I have two arrays "F1" and "F2":

F1 <- c(1,3,4,5)
F2 <- c(6,7,8,9)

I want to find nearest neighbor of each element in both "F1" and "F2".

Example: for value 3 in "F1", the nearest neighbor in "F1" is 4. And the nearest neighbor of 1 in "F2" is 6. And for value 9 in "F2", the nearest neighbor is 8, and 5 in "F1".

Answer 1

you could use the RANN package:

F1 <- c(1,3,4,5)
F2 <- c(6,7,8,9)
require(RANN)
nn2(F1,F2,k = 1)

which should produce the following results:

> nn2(F1,F2,k = 1)
$nn.idx
     [,1]
[1,]    4
[2,]    4
[3,]    4
[4,]    4

$nn.dists
     [,1]
[1,]    1
[2,]    2
[3,]    3
[4,]    4

Naturally, you may consider modifying the nn2 syntax to get more desirable results. The method uses the ANN library which gives you a lot of opportunities on how to find your nearest neighbours.

I'm thinking that you may be actually interested in finding the closest match . For example to find the closest match for the number 6 you could do:

which(abs(F1-6)==min(abs(F1-6)))

This is not the same as identifying the nearest neighbours.