Include SNN graphs to multiSVD — compute

Computing and including the shared nearest-neighbor (SNN) graphs for each modality and the target common manifold for an existing multiSVD object

Usage

compute_snns(
  input_obj,
  latent_k,
  num_neigh,
  bool_cosine = T,
  bool_intersect = T,
  min_deg = 1,
  tol = 1e-04,
  verbose = 0
)

input_obj: multiSVD class, after creation via create_multiSVD() or form_metacells()
latent_k: number of latent dimensions for the graph Laplacian bases
num_neigh: number of neighbors for each cell, when constructing the SNN graphs
bool_cosine: boolean, for using cosine distance if T or Euclidean distance if F
bool_intersect: boolean, on whether or not to symmetrize (via the AND function) the SNN
min_deg: minimum degree of each cell in the SNN
tol: small positive number
verbose: non-negative integer

updated multiSVD object