Run Diffusion Map on a precomputed distance matrix

diffusion.map(dist.mat, K = 10, sigma = NULL, nEV = 30, t = 1)

Arguments

dist.mat

matrix; Precomputed distance matrix (symmetric)

K

integer; Adaptive kernel bandwidth for each point set to be the distance to its K-th nearest neighbor.

sigma

numeric; Fixed kernel bandwidth, sigma will be ignored if K is specified.

nEV

integer; Number of leading eigenvectors to export

t

integer; Number of diffusion times

Value

List with the following elements:

diffu.embDiffusion Map embedding with the leading K eigenvectors
eigen.valsEigenvalues associated with corresponding eigenvectors