Create a K-nearest neighbor graph from data x. Uses
`nn2`

as a fast way to find the neares neighbors.

## Arguments

- x
data, a matrix, observations in rows, dimensions in columns

- k
the number of nearest neighbors.

- eps
number, if

`eps > 0`

the KNN search is approximate, see`nn2`

- diag
logical, if

`TRUE`

every edge of the returned graph will have an edge with weight`0`

to itself.

## Value

an object of type `igraph`

with edge
weight being the distances.