LaplacianEigenmaps: Laplacian Eigenmaps
In patrickreidy/phoneigen: Phonetic Analyses with Laplacian Eigenmaps

Compute the Laplacian-Eigenmaps embedding of a weighted adjacency matrix.

1	LaplacianEigenmaps(w, symmetric = FALSE)

`w`	A symmetric weighted adjacency matrix.
`symmetric`	A logical. If `FALSE`, then `w` is checked for symmetry. If it is known beforehand that `w` is symmetric, then it is faster to call with `symmetric = TRUE`.

The eigenvalues lambda and corresponding eigenvectors e that are solutions to the generalized eigenvector problem: L(w)*f = lambda*D(w)*e.

The eigenvalues and eigenvectors are returned as a list with two elements:

values: a numeric vector of the eigenvalues, in increasing order.
vectors: a tibble with columns x, e1, ..., eN, where N is the number of rows in the input matrix w. The column x contains the row names of w. The columns e1, ..., eN are the eigenvectors associated with the eigenvalues (e.g., vectors$e1 is associated with values[1]). Columns e2, ..., eJ give an embedding, in J-1-dimensinal space, for the data identified by x.