Fast Calculate Score Profile for Large Data

Calculates the multi-scale KL divergence score profile for large datasets with dyadic time scales (0, 2, 4, ...).

fast_calculate_score_profile_largeData(
  W,
  max_time = 2^15,
  init_state,
  score_ls = c("score0")
)

Arguments

W

matrix; the affinity matrix.

max_time

integer; the maximum diffusion time. Default is 2^15. The actual maximum diffusion time may be shorter if all nodes converge beforehand.

init_state

matrix; the row-wise normalized initial state matrix.

score_ls

character vector; the list of scores to compute. See get_score_profile for details. Default is c("score0").

Value

A data frame containing the computed scores for different diffusion times.